Strengths and Silences the Experiences of Lesbian, Gay, Bisexual and Transgender Students in Rural and Small Town Schools

Home , GLSEN

Strengths and Silences The Experiences of Lesbian, Gay, Bisexual and Transgender Students in Rural and Small Town Schools

A Report from the Gay, Lesbian & Straight Education Network www.glsen.org

Strengths and Silences The Experiences of Lesbian, Gay, Bisexual and Transgender Students in Rural and Small Town Schools

By: Neal A. Palmer, M.S. Joseph G. Kosciw, Ph.D. Mark J. Bartkiewicz, M.S.

A Report from the Gay, Lesbian & Straight Education Network www.glsen.org National Headquarters 90 Broad Street, 2nd floor New York, NY 10004 Ph: 212.727.0135 Fax: 212.727.0254

DC Policy Office 1012 14th Street, NW, Suite 1105 Washington, DC 20005 Ph: 202.347.7780 Fax: 202.347.7781 [email protected] www.glsen.org

ISBN 978-193409211-8

When referencing this document, we recommend the following citation:

Palmer, N. A., Kosciw, J. G., & Bartkiewicz, M. J. (2012). Strengths and Silences: The Experiences of Lesbian, Gay, Bisexual and Transgender Students in Rural and Small Town Schools. New York: GLSEN.

The Gay, Lesbian & Straight Education Network is the leading national education organization focused on ensuring safe schools for all lesbian, gay, bisexual and transgender students. Established nationally in 1995, GLSEN envisions a world in which every child learns to respect and accept all people, regardless of sexual orientation or gender identity/expression.

Electronic versions of this report and all other GLSEN research reports are available at www.glsen.org/research. Follow GLSEN’s Research Department on Twitter @GLSENResearch. Table of Contents

Preface...... vii

Acknowledgments...... viii

Executive Summary...... ix

Methods...... ix

Key Findings...... x

Biased Language in School...... x Overall Safety in School...... x Harassment & Assault...... x Educational Outcomes...... xi School Engagement...... xi Resources & Supports...... xii Conclusion & Recommendations...... xiii

Introduction...... 1

Methods...... 3

Results ...... 5

Biased Language in School...... 5 Overall Safety in School...... 8 Harassment and Assault...... 10 The Impact of Victimization on Educational Outcomes...... 14 School Engagement...... 17 Resources & Supports...... 23 The Benefits of LGBT-Related Resources & Supports...... 27

Conclusion & Recommendations...... 29

Limitations...... 29 Discussion...... 29 Future Directions for Research...... 30 Recommendations...... 31

Notes & References...... 33

Strengths and Silences iii

Tables & Figures

Table 1 Demographic Characteristics of Rural Survey Respondents...... 4 Table 2 School Characteristics of Rural Survey Respondents...... 4 Figure 1 Rural Students’ Frequency of Hearing Biased Remarks from Students...... 5 Figure 2 Rural Students’ Frequency of Hearing Biased Remarks from Staff...... 6 Figure 3 Frequency of Hearing Biased Remarks from Students by Locale...... 6 Figure 4 Intervention When Biased Remarks Were Made in Rural Schools...... 7 Figure 5 Feeling Unsafe at School by Locale...... 8 Figure 6 Avoiding Spaces at School by Locale...... 9 Figure 7 Feeling Unsafe at School Based on Sexual Orientation, by Locale and Region...... 9 Figure 8 Rural Students’ Frequency of Verbal Harassment in the Past School Year...... 10 Figure 9 Rural Students’ Frequency of Physical Harassment in the Past School Year...... 11 Figure 10 Rural Students’ Frequency of Physical Assault in the Past School Year...... 11 Figure 11 Experiences of Victimization, by Locale...... 12 Figure 12 Other Types of Harassment by Locale...... 13 Figure 13 Rural Students’ Frequency of Reporting Incidents of Harassment and Assault to School Staff...... 13 Figure 14 Effectiveness of Rural Students’ Reporting Incidents of Harassment and Assault to School Staff...... 13 Figure 15 Rural Students’ Absenteeism Due to Safety Concerns by Severity of Verbal Harassment...... 14 Figure 16 Rural Students’ Academic Achievement by Severity of Verbal Harassment...... 14 Figure 17 Rural Students’ Educational Aspirations by Severity of Physical Assault...... 15 Table 3 Discriminatory Policies and Practices Reported by Rural Students...... 16 Figure 18 Rural Students’ Outness to Peers and School Staff...... 18 Figure 19 Outness to at Least One Person by Locale...... 18 Figure 20 Outness to Students at School and Experiences of Victimization Based on Sexual Orientation, by Locale...... 19 Figure 21 Outness to Students at School and Self-Esteem, by Locale...... 19 Figure 22 Outness to Students at School and Depression, by Locale...... 19 Figure 23 Outness to Students at School and Sense of School Belonging, by Locale...... 19 Figure 24 Comfort Level Raising LGBT Issues in Class by Locale...... 20 Figure 25 Comfort Talking with School Staff about LGBT Issues by Locale...... 21 Figure 26 Talking to School Staff about LGBT Issues by Locale...... 22 Figure 27 Rural Students’ Sense of School Belonging and Talking about LGBT Issues to School Staff in the Past Year...... 22 Figure 28 Prevalence of LGBT-Related Resources in Schools by Locale...... 23 Figure 29 Sexual Health Curricula in Rural Schools...... 24 Figure 30 Frequency of GSA Attendance by Locale...... 24 Figure 31 Prevalence of LGBT Community Groups by Locale and Presence of a GSA...... 25 Figure 32 Number of Teachers and School Staff Supportive of LGBT Students by Locale...... 25 Figure 33 Acceptance/Support from Students and School Administration in Rural Schools...... 26 Figure 34 Number of Other LGBT Students at School by Locale...... 26 Figure 35 Type of Anti-Bullying and Harassment Policy by Locale...... 27 Figure 36 Rural Students’ Experiences of Victimization Based on Sexual Orientation and Presence of LGBT‑Related Resources...... 28 Figure 37 Rural Students’ Sense of School Belonging and Presence of LGBT-Related Resources...... 28

Strengths and Silences v

Preface

This past spring, Lenoir City High School, a school in rural Tennessee, featured a section of short articles on student life in its yearbook. One article, entitled “It’s OK to be Gay,” profiled openly gay student Zac Mitchell. Apparently, as far as Lenoir City School Board is concerned, it is not OK to be gay — or at least not OK to talk about it. Although the yearbook’s student staff and faculty advisors felt the article was a perfectly legitimate and inclusive depiction of life at the school, members of the school board were vocal in their opposition. “I don’t think that that type of material has any place in a yearbook,” said Board member Glenn McNish. Board Vice Chairman Rick Chadwick added, “It should not have been put in the yearbook, and it split our community, and we are going to straighten it out.”

The story highlights the challenges facing many LGBT students who attend schools in rural and small town areas, but also points to a resiliency and determination to use the resources available to them to make their schools safer for everyone. It is this complex reality that we see reflected in GLSEN’s new report on rural and small town LGBT students.

This new report from GLSEN, Strengths and Silences: The Experiences of Lesbian, Gay, Bisexual and Transgender Students in Rural and Small Town Schools, is the most recent release from our twelve years of research into the LGBT student experience nationwide, updated every two years via GLSEN’s National School Climate Survey. Our reports consistently illustrate the difficult learning environments encountered by LGBT students, and the discrimination, bullying and violence they experience in K-12 schools. The wealth of data we have collected also allows us to examine the experiences of specific segments of this student population, including transgender students (Harsh Realities) and LGBT students of color (Shared Differences). Whereas this report reveals numerous commonalities in experience among LGBT students across geographies, it also highlights the heightened victimization and lower access to resources experienced by rural and small town LGBT students as a result of physical and demographic isolation.

At every turn, research on the LGBT student experience represents both an urgent call to action and a roadmap for targeted advocacy, program development and service delivery. Strengths and Silences underscores the need for educators and policymakers to do more to address the safety risks for LGBT students in rural and small town schools. Rural LGBT students are far less likely to have access to LGBT‑related resources at school. Nonetheless, they benefit substantially when such resources are present. We must continue to do all we can to bring those critical in-school supports to every community in the country. As familiar as this call may now be, it will continue until all students, in every type of school and of every demographic, have access to the school-based supports that can transform the LGBT student experience and enable every student to thrive.

Eliza Byard, Ph.D. Executive Director GLSEN

Strengths and Silences vii Acknowledgments

The authors wish to thank the youth who participated in the survey and the LGBT youth services and programs that permitted their constituents to participate in the survey, as well as the organizations that assisted with disseminating information about the survey. The authors would also like to thank Alex Kowal for his work in the design of this report, and Ryan M. Kull, GLSEN Research Consultant, for his assistance with editing and proofing this report.

viii Strengths and Silences Executive Summary

For more than 20 years, GLSEN has worked to make schools safer for all students; it has sought Methods specifically to reduce the bullying and harassment targeted at students’ sexual orientation, gender Data used in this report come from the sixth identity, and gender expression. For lesbian, gay, installment of GLSEN’s National School Climate bisexual, and transgender (LGBT) students across Survey, which was conducted during the 2010– the country, violence and harassment experienced 2011 school year. GLSEN used two methods to in school affect their ability to learn. Although obtain a representative national sample of LGBT schools in urban areas are typically regarded as youth to participate in the survey: 1) outreach more violent or dangerous than schools in other through national, regional, and local organizations areas, findings from our National School Climate that provide services to or advocate on behalf of Surveys consistently show that it is most often rural LGBT youth, and 2) targeted advertising on the schools that may pose the greatest threats for LGBT social networking site Facebook. For the first students. It may be that community characteristics, method, we asked organizations to direct youth such as religious and cultural traditions, income, to the National School Climate Survey, which and educational levels, influence individual beliefs was available on GLSEN’s website, through their and attitudes toward LGBT people in these areas. organizations’ emails, listservs, websites, and social It may also be that a lack of positive LGBT-related networking sites. Additionally, a paper version of school resources negatively affects LGBT students’ the survey was made available to local community school engagement and academic performance, groups/organizations with limited capacity to particularly if they also experience bullying and access the Internet. To ensure representation of harassment. transgender youth, youth of color, and youth in rural communities, we made special efforts to notify Although research on the educational experiences of groups and organizations that work predominantly LGBT youth has grown considerably over the past 25 with these populations. For the second method, we years, less is known about rural students specifically. posted advertisements for the survey on Facebook, This research report examines the experiences of targeting all users between 13 and 18 years of age LGBT students in small town and rural areas on who gave some indication on their profile that they matters related to biased language in schools, school were lesbian, gay, bisexual, or transgender. safety, harassment and victimization, educational outcomes, school engagement, and LGBT-related When examining differences between rural students resources and support. It also examines the and suburban and urban students, we draw from prevalence and utility of LGBT-related resources the full sample of 8,584 LGBT secondary school in rural schools. Finally, this report concludes by students, specifically the 8,158 students for whom advocating for more intentional policies, measures, we collected reliable locale information. This report and programs that protect LGBT students. also examines in greater detail the experiences of the 2,387 students in the survey who attended schools in rural areas. These rural LGBT students were between 13 and 20 years of age, and most were White (78%) and identified as gay or lesbian (64%).

Strengths and Silences ix Key Findings Overall Safety in School Biased Language in School Most rural LGBT students in this survey felt unsafe in the past year due to personal characteristics like sexual orientation, gender, gender expression, race, Nearly all LGBT students in rural areas have heard disability, or religion. homophobic, racist, sexist, and negative gender expression-based remarks. Furthermore, students in • 81% of rural LGBT students had felt unsafe at rural areas more frequently experienced derogatory school during the past year because of a personal comments than students in suburban and urban characteristic. Sexual orientation and gender schools. For example: expression were the most common reasons rural students said they felt unsafe. • 97% of rural LGBT students heard “gay” used in a negative way (e.g., “that’s so gay”) sometimes, • Rural students were more likely than suburban often, or frequently in school. 94% heard other and urban students to feel unsafe in school, and homophobic language (“dyke” or “faggot”) rural students in the South and Midwest were sometimes, often, or frequently. more likely to feel unsafe than rural students in the West and Northeast. • 86% heard comments from students about someone not acting “masculine” enough sometimes, often, or frequently, and 69% Harassment and Assault heard such comments about students not acting “feminine” enough sometimes, often, or A majority of rural LGBT students had been verbally frequently. harassed because of their sexual orientation or gender expression, and substantial numbers had • A quarter or more of students also had heard experienced more severe physical harassment and school staff make homophobic remarks (25%), physical assault because of these characteristics. sexist remarks (30%), or negative remarks about Furthermore, although LGBT students across someone’s gender expression (35%) sometimes, the country experienced harassment and abuse, often, or frequently. students in rural areas were more frequently victimized than students in suburban and urban • Rural students were more likely than suburban areas. or urban students to hear most types of biased language, including homophobic remarks and • Nearly nine in ten (87%) rural LGBT students negative comments about gender expression. had been verbally harassed (e.g., called names or threatened) at school at least once in the past Rural LGBT students reported that school staff year on the basis of their sexual orientation, and members and students rarely intervened when 68% had been verbally harassed due to their biased comments were heard. gender expression. • Only 13% of rural LGBT students said staff • Nearly half (45%) of students had been members intervened most or all of the time when physically harassed (e.g., pushed or shoved) at homophobic comments were made, and only school at least once in the past year due to their 11% said that staff members intervened most sexual orientation. In addition, one-third (31%) or all of the time when negative comments were had been physically harassed because of their made about gender expression. gender expression. • Only 6% of students said that other students • One in five (22%) rural students said they had intervened most of the time or always when they been physically assaulted at school because of heard homophobic remarks, and 5% said that their sexual orientation in the past year, and 16% about comments regarding gender expression. said that they had been physically assaulted because of their gender expression. • Students in rural schools reported lower student and school staff intervention in homophobic remarks than suburban students.

x Strengths and Silences • Rural students experienced higher levels of • Rural LGBT students who experienced a victimization due to their sexual orientation and high severity of verbal harassment (occurring gender expression than suburban and urban frequently or often) based on their sexual students. orientation were significantly more likely to have missed school in the past month because • 70% of rural LGBT students had regularly had of feeling unsafe than students who were less rumors or lies spread about them, significantly severely verbally victimized (occurring never, more than suburban or urban students (61% and rarely, or sometimes): 53% of highly victimized 58%, respectively). students had missed school, compared to 29% • Two in five (40%) had regularly experienced of less severely victimized students. cyberbullying in the past year. Rural students • Rural LGBT students who had experienced a were more likely to have experienced high severity of verbal harassment based on cyberbullying than suburban and urban students. their gender expression were significantly more likely to miss school than students who had Intervention by teachers and other school staff experienced a lower severity of verbal harassment regarding incidents of harassment and assault can (54% compared to 23%). improve the school climate for LGBT students. Unfortunately, most rural LGBT students reported • Rural LGBT students who had experienced that such incidents were not effectively addressed by high levels of harassment and assault reported the staff in their schools. significantly lower grade point averages (GPAs) and college aspirations compared to students • Six in ten (60%) rural LGBT students said they who had experienced lower levels of victimization never reported incidents of harassment and (3.2 versus 2.9, for sexual orientation as well as assault to school staff or family members. Rural for gender expression). students, however, did not differ from urban or suburban students in the frequency of reporting School Engagement harassment and assault. For any LGBT student, experiences of harassment • When rural students did report incidents to and assault can negatively affect participation in school staff, two-thirds (68%) labeled the school activities and connectedness to school. LGBT responses as ineffective. Rural students rated students who are more out to peers and staff may staff responses to reporting as less effective than be more engaged in school but also experience more urban and suburban students. frequent victimization.

Educational Outcomes • Rural LGBT students felt less connected to their schools than suburban and urban students. In general, LGBT students, regardless of locale, often seek to avoid a hostile learning environment • Nearly half (45%) of rural students were by skipping classes or missing days of school. LGBT uncomfortable raising LGBT issues in class, students in rural areas, however, were slightly more similar to urban and suburban students. likely to miss classes or school for safety reasons than urban and suburban rural students. Also as • Rural students also felt uncomfortable talking we have found with LGBT students in general, rural with school staff about LGBT issues. Of all school LGBT students who were more severely victimized personnel, rural students were most comfortable missed even more classes or days of school. talking with teachers and counselors about LGBT issues. However, only about half said • One-third (36%) of rural LGBT students had they were comfortable doing so. Rural students missed days of school because they felt unsafe, were least comfortable talking with athletic greater than the 30% of suburban students and personnel, principals and vice principals, and 30% of urban students. security personnel: three-quarters or more of rural students said they were uncomfortable talking with these school personnel about LGBT issues.

Strengths and Silences xi • Rural students were less comfortable talking to • For those students who had a GSA at school, teachers, counselors, principals, and most other rural students more frequently attended GSA school staff members about LGBT issues than meetings than urban and suburban students. suburban or urban students. • Most rural students (94%) knew of at least • However, rural youth did not differ from suburban one teacher or staff person supportive of LGBT and urban students in their likelihood of talking students. However, fewer than half (42%) with school staff about LGBT issues. reported knowing 6 or more supportive staff members, which was lower than the 60% Being out in school can make LGBT youth more reported for suburban and 61% reported for engaged in the school setting. Therefore, outness is urban areas. a key indicator of school climate for LGBT students. • Only 5% of rural students attended schools • Rural students were out to peers, staff members, with comprehensive harassment and assault and parents at rates that were not different from policies (i.e., which enumerate protection on suburban and urban students. the basis of both sexual orientation and gender expression); one in five (19%) attended schools • For all students, being out in school was related with no bullying policies of any kind. Rural to higher rates of victimization. However, for rural students were least likely to attend schools with a students, being out was associated with even comprehensive or a partly enumerated policy. higher levels of victimization compared to urban and suburban students. Although they are less prevalent in rural areas, LGBT-related resources can make a significant Resources and Supports difference in the school environment for rural LGBT students. LGBT-related resources can help counter the negative effects of hostile school climates and serve • Rural students whose schools had LGBT-related as important tools in changing attitudes about LGBT supports such as GSAs, many supportive people. Given that students in rural schools had the school personnel, inclusive curricula, and highest incidence of victimization, they might be comprehensive anti-bullying policies, reported in greatest need of these supports. However, rural significantly lower levels of victimization due to LGBT students consistently reported less access to their sexual orientation and gender expression. LGBT-related support. • Rural students whose schools had LGBT-related • Only 11% of rural LGBT students reported that resources also reported higher levels of school their curricula included information on LGBT belonging, higher self-esteem, and lower levels of people, history, or events (compared to 18% depression. of suburban and 20% of urban students), and only 13% said their textbooks included such information (compared to 20% of suburban and urban students). • Over a third (39%) of rural LGBT students reported that they could access LGBT-related information through school computers, which was lower than that for suburban and urban students (44% for both). • Rural students were half as likely to have a GSA (Gay-Straight Alliance) or other student club that addresses LGBT issues as suburban and urban students (27% of rural students vs. 55% of suburban and 53% of urban students).

xii Strengths and Silences Conclusion and Recommendations

Findings from this report demonstrate that LGBT students across the country — particularly those in rural and small town areas — encounter hostile school environments. Compared to students in urban and suburban areas, LGBT students in rural schools are more likely to hear negative comments about gender expression and sexual orientation; feel unsafe at their schools due to their sexual orientation, gender identity, or gender expression; and experience verbal and physical harassment and assault due to these characteristics. In addition, the more hostile climates experienced by students in rural and small town schools may be further exacerbated by the lack LGBT‑related resources relative to their suburban and urban counterparts, including a lower prevalence of GSAs, supportive staff, inclusive curricula, and comprehensive anti‑bullying policies.

These findings demonstrate a clear need for safer and more inclusive learning environments for LGBT students in rural and small town areas. Educators, policymakers, and supporters of safe school initiatives can use the information from this report to better understand the specific experiences of rural LGBT students and take appropriate steps to make rural schools safer and more inclusive for LGBT students. Developing LGBT-related resources and supports may require additional support or alternative strategies in rural areas. Advocates should strive to expand use of the resources that already exist. For instance, for rural students who may be living in vast geographic areas, online resources may be useful to supplement or make up for a lack of resources. Community members may find it valuable to establish community groups and programming for LGBT youth, as they seem to be particularly absent from rural areas. Cultivating more intentional involvement from educators may also be helpful, given that supportive educators are associated with academic and psychological benefits regardless of locale. We recommend that educators, policymakers, and supportive community members begin by organizing community-based coalitions of individuals and organizations that concern themselves with school safety in order to build broader community support for LGBT‑specific policies and practices. Together, these recommendations will help make schools safer for all students in school, regardless of sexual orientation, gender identity, gender expression, or locale.

Strengths and Silences xiii

Introduction

For more than 20 years, GLSEN has worked to educational experiences of LGBT youth has grown make schools safer for all students; it has sought considerably over the past 25 years, less is known specifically to reduce the bullying and harassment about rural students specifically. Previous GLSEN targeted at students’ sexual orientation, gender research has shown that even in elementary school, identity, and gender expression. For lesbian, gay, teachers in rural areas are less likely to address bisexual, and transgender (LGBT) students across remarks by students that use “gay” in a negative the country, violence and harassment experienced way as compared to teachers in suburban and in school affect their ability to learn. Although urban urban areas, and are also less likely to believe that schools are typically regarded as more dangerous or students who may grow up to be LGBT would feel violent than rural or suburban schools1, it may be comfortable at their schools.8 Gay‑Straight Alliances rural school settings that pose the greatest danger (GSAs) have been shown to develop more slowly for LGBT students.2 in rural areas than in urban areas.9 In addition, school personnel may be ill‑equipped to respond to LGBT people in rural areas, regardless of age, LGBT‑related harassment in schools, whether this may have more negative experiences related to inaction is intentional or not.10 Furthermore, there is their LGBT identity than those in other areas. The some evidence from prior research that rural LGBT stigmatization of the larger LGBT population in rural youth also typically lack the support of more formal areas is reflected in informal beliefs about LGBT institutions that help buffer against experiences of people as well as in the absence of more formal victimization in urban areas, such as bookstores, protections and resources for LGBT people. Adults coffee shops, community organizations, and in rural areas are more likely to have unfavorable LGBT‑focused school or workplace social groups or opinions of gay men and lesbians and be more organizations.11 However, research also suggests that uncomfortable around them, and also more likely to LGBT youth in rural areas adapt and make use of the oppose same‑sex marriage, compared to residents resources that are available in their communities. of other parts of the US.3 In addition, evangelical For example, rural youth may meet and gather in Christianity, lower income, and lower adult education local parks or shopping center parking lots in the levels, all of which are more prevalent in rural absence of LGBT community centers.12 areas4, tend to be associated with more conservative social beliefs, including opposition to same‑sex This research report takes an in‑depth look at the marriage.5 Unsurprisingly, and perhaps as a result, experiences of LGBT students in rural and small rural areas are less likely to have LGBT institutional town areas on issues such as biased language in protections, as seen in the lower prevalence of schools, school safety, harassment and victimization, inclusive sexual orientation and gender expression educational outcomes, school engagement, non‑discrimination ordinances in rural areas than in and LGBT‑related resources and support. Given suburban or urban areas.6 that previous research has found that there are regional differences in attitudes toward LGBT Negative attitudes and a lack of protections in the people — specifically, that places in the South and larger community might be expected to be reflected Midwest may be more hostile than areas of the West in the school community as well. For LGBT youth and Northeast — we also examine the influence specifically, this general negative climate in rural of geographic region in the experiences of rural areas may contribute to a lower prevalence of LGBT students.13 Finally, this report concludes by resources that could be useful to LGBT youth, and advocating for more intentional policies, measures, thus result in more hostile school climates for LGBT and programs that protect LGBT students, students in these areas.7 Although research on the specifically those in rural areas.

Strengths and Silences 1

Methods

Data used in this report come from the 2011 outreach efforts were made to notify groups and installment of GLSEN’s National School Climate organizations that work predominantly with these Survey, which is a biennial survey of U.S. secondary populations about the survey. school students who identify as lesbian, gay, bisexual, and/or transgender. Youth were eligible Contacting participants only through LGBT to participate in the survey if they were at least youth‑serving groups and organizations would have 13 years of age, attended a K–12 school in the limited our ability to reach LGBT students who United States during the 2010–11 school year, were not connected to LGBT communities in some and identified as lesbian, gay, bisexual, or a sexual way. Thus, in order to broaden our reach to LGBT orientation other than heterosexual (e.g., queer, students who may not have had such connections, questioning) or identified as transgender or as we conducted targeted advertising on Facebook. having a gender identity other than male, female, or Notices about the survey were shown to users transgender (e.g., genderqueer). In order to obtain between 13 and 18 years of age who gave some a more representative sample of LGBT youth, two indication on their profile that they were lesbian, methods were used to locate possible participants. gay, bisexual, or transgender.

First, the National School Climate Survey was made The survey collected information from 8,584 LGBT available online through GLSEN’s website. Notices students, though only the 8,158 students for whom about the survey were sent through GLSEN’s email we collected reliable locale information are used for and chapter networks, as well as through national, analysis in this report.14 Students were from all 50 regional, and local organizations that provide states and the District of Columbia and from 3,224 services to or advocate on behalf of LGBT youth. unique school districts. Rural students comprise National and regional organizations posted notices 29.4% (n=2,387) of the full sample of the 2011 about the survey on listservs, websites, and social survey.15 Table 1 presents the rural subsample’s networking websites (e.g., TrevorSpace). Local demographic characteristics and Table 2 shows the community groups serving LGBT youth notified their characteristics of the schools they attended. About participants about the online survey via email, social four‑fifths of the rural subsample (78%) were White, networking, and flyers. In addition, a paper version half (48%) were female, and two‑thirds identified as of the survey was made available to local community gay or lesbian (64%). Students were in the 6th to groups with limited capacity to access the Internet 12th grades, with the largest numbers in 10th and (resulting in 139 completed paper surveys). To 11th grades. Compared to the suburban and urban ensure representation of transgender youth, youth respondents, rural LGBT students were more likely of color, and youth in rural communities, special to be White, attend a public school, live in the South and Midwest, and attend smaller schools.16

Strengths and Silences 3 Table 1. Demographic Characteristics of Rural Survey Table 2. School Characteristics of Rural Survey Respondents (N=2387) Respondents (N=2387) Gender Identity† Region Female 48.0% n=1145 Northeast 19.6% n=465

Male 37.6% n=895 South 36.5% n=865

Transgender 7.8% n=185 Midwest 31.0% n=735

Other 6.6% n=158 West 12.9% n=307 Sexual Orientation School Type Gay/Lesbian 63.8% n=1522 Public 95.6% n=2214

Bisexual 26.7% n=637 Magnet 2.0% n=44

Other Sexual Orientation 9.6% n=228 Charter 3.1% n=68 (e.g., Queer, Questioning) Religious-Affiliated School 1.3% n=31 Race‡ Other Independent or Asian or Pacific Islander 0.7% n=16 Private School 3.1% n=71

Native American, American 0.8% n=19 School Size Indian, or Alaska Native Small (<500 students) 27.6% n=654 White or European 78.3% n=1860 American Medium (501‑1000) 34.7% n=823

African American or Black 1.9% n=44 Large (1001‑1500) 23.0% n=544

Middle Eastern or Arab 0.9% n=21 Very Large (>1500) 14.7% n=348 American, any race

Hispanic or Latino/a, 8.6% n=204 any race

Multiracial 8.8% n=210 Grade 6th 0.2% n=5

7th 2.5% n=59

8th 9.0% n=214

9th 18.0% n=429

10th 24.9% n=592

11th 24.0% n=571

12th 21.4% n=508 Age (mean) 16.10 years

† “Female” includes participants who selected only female as their gender, and “male” includes participants who selected only male. The category “transgender” includes participants who selected transgender, male-to-female, or female-to-male as their gender, including those who selected more than one of these categories. Participants who selected both male and female were categorized as “other” (e.g., genderqueer, androgynous).

‡ Participants who selected more than one category were coded as “multiracial”, with the exception of participants who selected “Hispanic or Latino” or “Middle Eastern or Arab American”.

4 Strengths and Silences Results

Students’ Reports of Hearing Biased Language in School Biased Remarks in School One of the more pervasive negative elements of Biased remarks were commonly heard by LGBT the school climate is biased language. Students students in rural schools. As shown in Figure 1, frequently hear negative comments on the basis nearly all students had heard biased remarks in of personal attributes — including race, sexual school based on gender, gender expression, sexual orientation, gender identity, gender expression, and orientation, and other personal characteristics. More religion — whether they are directed at individual than 98% had heard sexist remarks, “gay” used students or expressed more generally. in a negative way, or other homophobic remarks at school, and more than 90% had heard racist The 2011 National School Climate Survey asked remarks or negative remarks related to gender students about the frequency at which they hear expression. biased remarks, such as those that are homophobic, racist, or sexist in nature. Students were asked about The most prominent derogatory remark was “gay” hearing these remarks from other students as well as used in a negative way, such as “that’s so gay” from school staff, and were also asked how students to refer to something inferior, undesirable, or less and staff intervened when these remarks were made. valuable. In fact, 69% of rural LGBT youth reported Although hearing homophobic, racist, sexist, and that they frequently heard “gay” used in this way; gender expression‑based remarks were prevalent only 9% reported that they heard such comments across locales, LGBT students in rural areas more only sometimes, rarely, or never. Other homophobic frequently heard derogatory comments than students comments were also commonly overheard in school. in suburban and urban schools.

Strengths and Silences 5 “No homo” is a more recent phrase used to rid make homophobic remarks (25%), sexist remarks a statement of a homosexual connotation. For (30%), or negative comments in regard to gender instance, someone might utter the phrase after a expression (35%) at least sometimes. Given that compliment to someone of the same gender, as in, school personnel have a direct responsibility in “I like your jeans — no homo.” Although it was not promoting a safe learning space for students, any heard as commonly as “gay” in a negative manner, use of such language is unacceptable and perhaps half (52 %) of rural LGBT students reported hearing reinforces a negative climate for students. the phrase frequently or often. Other homophobic remarks, such as use of the word “faggot” or Commonly hearing biased remarks in school was “dyke,” were heard regularly as well, with 79% not unique to students in rural areas. However, as of respondents reporting hearing such comments shown in Figure 3, rural students reported hearing frequently or often.

Rural LGBT students were also exposed to derogatory comments regarding sex and gender expression. More than half reported frequently overhearing sexist comments, such as use of “bitch.” Three in five (61%) students reported frequently or often hearing remarks about students not acting “masculine” enough; 42% had heard similar comments about students not acting “feminine” enough.

Frequencies of racist language were generally lower than homophobic and sexist remarks. Nevertheless, nearly half (44%) of respondents reported hearing racist remarks frequently or often.

Students were also exposed to biased language from their teachers and other school staff, as shown in Figure 2. Although school staff made biased remarks less regularly than students17, a quarter or more of students reported hearing a school staff member

6 Strengths and Silences most types of biased language more than urban and were much less likely to report that teachers and suburban students.18 The differences across locale school staff intervene frequently when hearing other were particularly strong for homophobic remarks, types of biased language.21 For instance, only 13% such as “fag” and “dyke.” The one exception is with of students said staff members intervene always the phrase “no homo” — rural students heard this or most of the time when they hear homophobic expression slightly less often than suburban or urban remarks, and only 11% said staff members students.19 This difference may reflect the origins of intervene always or most of the time when they the phrase in hip‑hop culture20, which tends to be hear negative remarks related to gender expression. associated more with urban (and perhaps suburban) Only a third of students (32%) said staff members places than rural ones. respond always or most of the time when they overhear sexist remarks in school. Differences by Intervention with Biased Language by locale for staff member response to biased remarks School Staff and Students were minimal, with the exception of homophobic remarks: staff members in suburban areas were The manner in which school staff respond to more likely to intervene when homophobic remarks biased language can also influence the overall were made than staff members in rural areas.22 school climate. If the use of biased language goes unchallenged in the school setting, then it can Although it is primarily teachers’ and other staff signal that such language is acceptable for use in members’ responsibility to intervene in instances the school and perhaps in other public spaces. On of biased language, students may, at times, the other hand, if staff members intervene when intervene as well. It is possible that intervention they hear such language, then they may be sending by students may have a greater impact, as it may a message that such language is unacceptable. model appropriate behavior for other students. As shown in Figure 4, biased language was largely Unfortunately, students were even less likely to unchecked in rural schools. Rural LGBT students intervene than staff members.23 For instance, only reported that teachers and other staff members more 18% of the rural LGBT students said other students frequently intervene when they hear racist remarks; intervene most of the time or always when they nevertheless, only a little more than half (54%) overhear racist language in school. Fewer than reported that educators intervene always or most of one in five (17%) said students intervene most the time with this type of remark. Rural students of the time or always when they overhear sexist remarks. Intervention by students was even lower for comments related to sexual orientation and gender expression.24 Only 6% and 5% of rural LGBT students indicated that other students in their school respond most of the time or always to homophobic remarks or negative remarks about gender expression. Differences by locale for student response to biased remarks were minimal, although students in rural and suburban areas were slightly less likely to intervene when homophobic remarks were made than students in urban areas.25 The failure of school staff and students to intervene, particularly when derogatory remarks about sexual orientation and gender expression are made, may contribute further to a hostile environment for LGBT students.

Strengths and Silences 7 more likely to feel unsafe on the basis of their sexual Overall Safety in School orientation (71% vs. 62% of suburban and 58% of urban school students)27 and gender expression Students who perceive themselves to be different (49% of rural students vs. 42% of suburban and from their classmates may feel less safe in school. 42% of urban students).28 The differences by Students in the survey were asked whether they locale were largest for feeling unsafe because of had felt unsafe in the past year due to personal sexual orientation.29 In addition, rural students were characteristics, such as sexual orientation, gender, more likely to feel unsafe due to their religion than gender expression, race, disability, or religion. suburban or urban students.30 Feeling unsafe due Furthermore, students were asked if there were to gender, race, or disability was less common for spaces they avoided in school for safety reasons. all students, regardless of locale. There were no As with biased language, rural students were significant differences by locale for feeling unsafe more likely to feel unsafe at school due to sexual due to gender or disability, though students in rural orientation or gender expression than suburban and areas were less likely to say they felt unsafe due to urban students. their race.31

Feeling Unsafe at School When asked about specific places in school they avoided because they felt unsafe, rural students An overwhelming majority (81%) of rural LGBT were more likely to report having avoided bathrooms students reported that they had felt unsafe at and locker rooms because of feeling unsafe or school during the past year due to their gender, uncomfortable than other spaces, such as the sexual orientation, race, religion, or other personal cafeteria, school buses, and hallways (see Figure characteristic, and this number was substantially 6).32 Although the same pattern applied to suburban more than the 73% of suburban students and 71% and urban students as well33, rural students were of urban students who reported feeling unsafe more likely to feel unsafe than suburban and 26 for any of these reasons. Rural LGBT students urban students in all of these spaces except for the most commonly felt unsafe due to their gender cafeteria and other, non‑specified school grounds expression and sexual orientation, as shown in and places.34 In particular, the largest differences Figure 5. Although this was also true for suburban across locale were observed for bathrooms and and urban students, rural LGBT students were even locker rooms.35

8 Strengths and Silences Prior research about geographic differences across the U.S. indicates that rural areas of the South and Midwest may be more socially conservative than other areas of the country.36 For example, a recent study of social and religious values found that 52% of Southerners and 45% of Midwesterners disapproved of same‑sex marriage, compared to 39% of Westerners and 35% of Northeasterners.37 In addition, in our 2011 National School Climate Survey, we found that LGBT students experienced more hostile climates in these regions. Thus, we wanted to examine how region might be related to rural students’ feelings of safety. As shown in Figure 7, across all regions of the country, students in rural areas felt less safe than students in suburban and urban areas.38 In addition, rural students in the South and Midwest were more likely to report feeling unsafe based on sexual orientation than were students in rural areas of the Northeast or West.39

Strengths and Silences 9 seen in Figure 8, nearly nine in ten (87%) rural Harassment and Assault students had been verbally harassed at least once in the past year at school on the basis of their sexual Other experiences in the school environment can orientation, with 43% reporting that the harassment also contribute to a hostile school climate for LGBT occurred frequently or often. In addition, 68% had students. In addition to overhearing biased language been verbally harassed at least once in the past year in school, LGBT students can be the direct targets due to their gender expression, with 29% reporting of verbal harassment and physical harassment and that it occurred frequently or often. assault. Students in this survey were asked about experiences of verbal and physical harassment and Physical Harassment physical assault related to such characteristics as sexual orientation, gender, gender expression, race, Students also reported being physically harassed disability, and religion. It is the responsibility of in the past school year because of personal school staff to address these instances of bullying characteristics, which includes being the victim of and harassment in school. For this reason, we asked pushing, shoving, or similar physical aggression. students about school staff responses to incidents of Although rates of physical harassment were lower harassment. than for verbal harassment, a substantial number of students had been physically harassed at some Verbal Harassment point in the preceding year at school. As shown in Figure 9, rural LGBT students were most commonly Rural LGBT students reported high rates of verbal physically harassed due to their sexual orientation42 harassment, or being the target of name‑calling and gender expression.43 Nearly half (45%) of or threats due to characteristics such as their students had been physically harassed at least once sexual orientation, gender, gender expression, in the past year due to their sexual orientation, with race, disability, or religion. As with hearing biased 16% saying that it occurred frequently or often. remarks, LGBT students in rural areas most In addition, one third (31%) had been physically frequently experienced harassment based on their harassed because of their gender expression, with 40 41 sexual orientation and gender expression. As 10% saying that it occurred frequently or often.

10 Strengths and Silences Physical Assault on the basis of their sexual orientation44 and gender expression.45 One‑fifth (22%) said they Rural LGBT students also experienced direct had been physically assaulted because of their physical assault because of their personal sexual orientation in the past year at school, with characteristics, which includes violent actions such 6% saying that it occurred frequently or often. In as being kicked, punched, or injured with a weapon. addition, 16% said that they had been physically As depicted in Figure 10, rural LGBT students most assaulted because of their gender expression, with commonly experienced this form of victimization 5% saying that it occurred frequently or often.

Strengths and Silences 11 Differences in Victimization by Locale had experienced them regularly (sometimes, often, or frequently). Three in four (78%) had regularly In addition to understanding the different types felt excluded or left out or had experienced rumors of victimization experienced by LGBT students in or lies being spread about them (70%). In addition, rural schools, we also wanted to examine whether almost half (44%) had regularly (sometimes, these rural students experienced higher levels of frequently, or often) experienced sexual harassment, victimization than their peers in other locales. The which includes having sexual remarks made results suggest that although LGBT students across toward them or having someone touch their body locales commonly experienced harassment and inappropriately. Finally, more than a third (40%) assault, the problem was more severe for LGBT had regularly experienced some form of electronic students in rural areas. As shown in Figure 11, harassment, or “cyberbullying” (e.g., being the rural students were found to experience greater target of negative comments or attacks on MySpace levels of victimization46 than both suburban and or Facebook; or receiving targeted, negative emails urban LGBT students on the basis of their sexual or text messages). orientation47, gender48, gender expression49, and religion.50 As with feeling unsafe at school, rural For each locale, having rumors or lies spread, students reported the largest differences between feeling excluded or left out, and sexual harassment urban and suburban students for victimization based were the most common other forms of harassment. on sexual orientation.51 Levels of victimization based Nonetheless, LGBT students in rural areas on disability were not different by locale52, and experienced each of these types of harassment rural and suburban students experienced slightly more frequently than students in suburban or less victimization based on race compared to urban urban areas.54 Again, these findings suggest that students.53 although experiences of harassment are common across locales, rural students often experience these The survey also asked students about other negative problems to a higher degree. events they may have experienced in school, such as being sexually harassed; having property damaged Reporting of Harassment and Assault or stolen; being the target of cyberbullying; or being the target of relational bullying (having rumors or lies Students were also asked about how often they spread about them or feeling intentionally “left out” reported incidents of harassment and assault to of some event, gathering, or other social activity). school staff and whether the school staff member’s Most rural LGBT students had experienced each of response was effective. When students fail to report these other types of harassment at least once in the such incidents, their experiences of victimization past year, and as Figure 12 shows, large numbers are likely to go unaddressed, and few efforts may

12 Strengths and Silences be directed to improving their school experience. If We also asked about the effectiveness of school students feel like staff members respond effectively staff responses when students did report instances to incidents of harassment, they may feel better of victimization. Among the rural LGBT students supported in school and have better overall school who reported incidents to school staff, only a third experiences. (32%) of students said that the response to their reports was somewhat or very effective, yet half Unfortunately, many students do not feel (49%) said that the responses by school staff were comfortable reporting experiences of harassment completely ineffective (see Figure 14). Given the low and assault. As shown in Figure 13, more than numbers of students who found staff intervention to half of students said they never reported these be at all effective, it is perhaps not surprising that incidents to school staff (60%). Furthermore, only students so infrequently reported their experiences 14% of rural students said they reported incidents of harassment and assault. Although rural students of harassment and assault to school staff always or did not differ from urban and suburban students, most of the time. Rates of reporting did not differ by they were significantly more likely to rate the staff locale.55 members’ responses ineffective.56

Strengths and Silences 13 times during the past month. Rural LGBT students The Impact of exhibited moderately higher rates of absenteeism than urban or suburban students.58 Compared to Victimization on the 36% of rural students who had missed a day of school in the past month, only 30% of suburban Educational Outcomes LGBT students and 30% of urban LGBT students had done so. Biased language, harassment, and assault may negatively affect LGBT students’ ability to receive The effect of harassment on educational outcomes an education. Being harassed may interfere with a appeared to be directly related to the severity of that student’s ability to concentrate during class or even harassment. Students who experienced high levels at home. In addition, students may attempt to avoid of victimization due to their sexual orientation or the spaces where experiences of victimization occur, gender expression were more likely to miss class or such as the classroom or bathroom. Higher rates of school than students who experienced lower levels absenteeism may, in turn, lead to poorer academic of victimization. For instance, as shown in Figure performance. In the 2011 National School Climate 15, rural LGBT students who experienced a high Survey, we found that higher rates of victimization severity of verbal harassment (occurring frequently were associated with lower grades, lower educational or often) based on their sexual orientation were aspirations, and greater absenteeism due to safety significantly more likely to miss some school in the 57 concerns for LGBT students across locales. In past month (53%) than students who experienced this report, we examine the relationships between less victimization (29%, occurring never, rarely, or victimization and educational outcomes for rural sometimes).59 LGBT students specifically. Rural students who had experienced high levels Absenteeism of victimization had significantly lower grade point averages (GPAs) than students who had experienced One third (36%) of rural LGBT students had missed lower levels of victimization, even after accounting a day of school in the past month due to feeling for school absences.60 For instance, as depicted in unsafe, including one in ten (12%) rural students Figure 16, students who had experienced high levels who said they had missed school four or more of verbal harassment due to their gender expression

14 Strengths and Silences discrimination in the school environment, whether resulting from formal school or district policies, or from informal practices by school personnel. In the 2011 National School Climate Survey, we asked students to describe ways they felt their schools discriminate against LGBT people. Rural students were more likely to report that their schools had discriminatory policies and practices than suburban and urban students (25% of rural students, compared to 20% of suburban and 20% of urban students).63

Although rural students were more likely to report that their schools had discriminatory policies and practices, the types of discrimination they experienced were similar across locales. Rural LGBT students most commonly reported that their schools discriminated against LGBT relationships (31.8% of students who reported discriminatory policies and practices). For instance, they said they were more severely punished for public displays of affection than non‑LGBT students, and a couple of students said that their privacy had been violated when their school disclosed their sexual orientation or had lower GPAs than students who had experienced transgender identity to their parents. Many students less severe levels of verbal harassment due to their mentioned that school functions discriminated sexual orientation, corresponding to the difference against LGBT couples: between a C and a B (2.9 vs. 3.2) in most schools. My school does not allow same‑sex couples Higher severities of physical harassment and to attend prom. They have to be checked in assault were also associated with negative academic as individuals while straight couples can be outcomes. As indicated in Figure 17, rural LGBT checked in as couples. (Female student, 11th students who experienced a high severity of physical grade, IL) assault (occurring frequently or often) based on 61 62 gender expression or sexual orientation were Rural LGBT students also commonly reported staff significantly less likely to plan to attend college than practices that promoted negative attitudes toward students who experienced less severe victimization LGBT people (22.5% of those who responded to (occurring never, rarely, or sometimes). For instance, the question), including the presence of anti‑LGBT only 84% of rural LGBT students who experienced content in class, non‑intervention in biased language a high severity of physical assault based on gender and harassment, and differential enforcement of expression planned to attend college, compared to disciplinary policies toward LGBT students. Several 93% of students who experienced a low severity of students commented that staff members themselves physical assault. Similar patterns were found for used biased language: physical harassment based on gender expression or sexual orientation. Many teachers reinforce the bullying by always suggesting that male students should Discriminatory Policies and Practices act a certain way; otherwise that student is gay. (Male student, 11th grade, MN) Hearing homophobic and negative remarks about gender expression in the hallways and directly Counselors often tell us that we bring experiencing victimization from other students harassment upon ourselves, and the term are overt contributions to a hostile climate for ‘gay’ in a demeaning manner is tossed around LGBT students. In addition, LGBT students may by faculty. (Transgender student, 12th grade, be negatively affected by less overt experiences of UT)

Strengths and Silences 15 One in five (21.5%) rural LGBT students who In addition, 14.5% of rural LGBT students who responded to the question reported that their said that their schools had discriminatory policies school limited their ability to discuss LGBT issues or practices reported that their schools reinforced in the school setting, including in the suppression gender boundaries around dress: of GSAs and similar clubs, limits on LGBT‑related discussions in the classroom, and discouragement Graduation: Girls wear a certain color and of staff support of LGBT issues. Many students said MUST be in a dress. Boys wear another color that they were prevented from expressing themselves and must be in pants with a tie. For band fully as LGBT individuals: concerts, boys and only boys get to wear bowties. Girls must be in feminine clothing. We’re not allowed to post any information (Other gender student, 12th grade, PA) on our GSA around the school or publicly announce anything GSA related. Everything has to be done in private; Facebook, talking to people privately. (Female student, 9th grade, MI)

Table 3. Discriminatory Policies and Practices Reported by Rural Students (N=600) (number reporting % of students response) Policies and Practices that Discriminate Against LGBT Relationships 31.8% (191) Rules Regarding Dances and School Functions 19.5% (117) Enforcement of Public Displays of Affection 14.7% (88) Violations of Student Privacy 0.3% (2) Policies and Practices that Reinforce Gender Boundaries around Dress 14.5% (87) Policies and Practices that Segregate School Activities Based on Gender 5.5% (33) Policies and Practices that Particularly Affect Transgender Students 4.7% (28) Gender-Segregated Locker Rooms and Gyms 4.3% (26) Use of Gendered Pronouns and Legal Sex 0.7% (4) Policies and Practices that Limit Discussion of LGBT Issues 21.5% (129) Suppression of GSA Efforts 7.3% (44) Restrictions on LGBT-Related Self Expression 13.2% (79) Limits on Discussion of LGBT Issues in the Class and School Activities 2.3% (14) Suppression of Staff Support for LGBT Students/Issues 0.8% (5) Staff Practices that Promote Negative Attitudes toward LGBT People 22.5% (135) Use of Biased Language 13.7% (82) Anti-LGBT Content in Classes 2.2% (13) Non-Intervention in Biased Language and Victimization 13.3% (80) Differential Enforcement toward LGBT Students 1.3% (8) Absence of Supportive Policies and Practices 12.8% (77) Lack of LGBT Curricular Content 5.5% (33) Lack of LGBT-Related School Resources 1.5% (9) Non-Inclusion in Bullying and Harassment Policies 6.0% (36) Other Discriminatory Experiences in Schools 4.2% (25)

16 Strengths and Silences Rural LGBT students also reported policies and practices that specifically affected transgender School Engagement students (4.7% of those who responded to the question), including in the refusal of school staff In addition to reducing academic achievement to use preferred gender pronouns. Others said that and educational aspirations, harassment can also students at their school were only permitted to use negatively affect how engaged students are in the bathrooms or locker rooms of their legal sex, school and with their peers at school. Given that which sometimes exposed transgender students in LGBT students are frequently targeted because particular to danger from other students or personal of personal characteristics, they might feel less discomfort: connected to their peers. For example, students may psychologically withdraw or feel emotionally Male/female locker rooms made for a difficult time. detached from their schools when they are targets of A trans friend of mine (female to male) was not harassment and they may be less engaged in class allowed to use the male locker rooms. Before this and in school activities. Considering that the school incident, no one knew he was biologically female. climate may be especially hostile in rural areas, He got made fun of mercilessly. (Female student, some students may not feel comfortable being out to 12th grade, NH) their peers, school staff, or members of their family. LGBT students may also feel obliged to keep silent Finally, some students (12.8% of those who about LGBT issues if they feel their opinions are not responded) commented that the absence of valued in the school setting. In order to examine inclusive and supportive policies and resources felt possible effects of a negative school climate on discriminatory, such as the lack of LGBT content student engagement in school, students in this in the curriculum and a lack of relevant resources survey were asked about their feelings of belonging in the school, including access to LGBT‑related at school; how out they were to peers and school information through school computers. Several personnel; and how frequently they raised LGBT students commented on the lack of inclusion of issues in class and spoke with school staff about sexual orientation and gender identity/expression in LGBT issues. their school’s anti‑bullying and harassment policy: School Belonging There’s really no specific policy that discriminates, but they don’t have anything to School belonging is an important indicator of the protect us, like an anti‑discrimination policy quality of LGBT students’ school experience because or anything. I was afraid to be an out lesbian it is related to educational outcomes: in general, here, but I am and I am getting hate for it. students who feel more connected to their school 64 (Female, 11th grade, AR) perform better academically. Unfortunately, LGBT students who are more victimized in school report 65 Together, these responses indicate that rural LGBT lower school belonging. In order to assess school students are exposed to policies and practices that belonging, students were asked how much they negatively affect their educational experiences agreed or disagreed with a series of statements in schools, as was found for students across the about their relationship to and connection with 66 country. As with the more overt biased language their schools. Rural LGBT students reported and victimization, such discriminatory policies and significantly lower levels of school belonging than practices may contribute negatively to the school either suburban or urban students (2.40 compared 67 climate. to 2.56 and 2.59). In that rural LGBT students experienced higher levels of victimization and were exposed to higher levels of biased remarks, it may have been the case that their decreased feelings of belonging were a result of higher victimization. However, even among LGBT students who reported fewer incidents of biased remarks and victimization, rural students reported feeling less connected to their schools than suburban or urban students.68 This finding suggests that additional factors — such as lower community support and fewer resources

Strengths and Silences 17 in school may contribute to rural LGBT students Being out may identify a student as a potential feeling less connected to their schools, a possibility target for harassment based on sexual orientation, that researchers and community advocates should gender expression, and other characteristics. In our examine further. 2011 National School Climate Survey, we found that for all LGBT students, being more out in schools Outness was related to increased victimization.72 Although rural students were not more or less likely to be Another indicator of how much an LGBT student out compared to their suburban and urban peers, feels like a part of their school community can be they were more likely to experience victimization. how open they are to their peers and school staff Thus, it is possible that they might face more severe about their sexual orientation or gender identity. consequences for being out. As shown in Figure 20, Despite the risks involved in coming out, including across all locales, students’ levels of victimization greater victimization and lower educational increased as their outness in school increased.73 outcomes, coming out is also traditionally associated However, the negative effect for being out was with some positive outcomes, such as improved much stronger in rural areas; students who were psychological well‑being.69 out to all or most of their peers were much more severely victimized in rural schools than students in As shown in Figure 18, LGBT students were much suburban or urban schools.74 more likely to be out at school to peers than to a staff person.70 Only 7% of rural LGBT students were As mentioned, LGBT students who are more not out to at least one other student, compared to open about their sexual orientation and/or gender one‑third (34%) of students who were not out to a identity at school may have more positive feelings single school staff person. about themselves and may feel a greater part of their school. As shown in Figure 21, rural LGBT Given the elevated levels of victimization observed students in this survey who were out to more peers in rural areas, it is perhaps surprising that rural in school reported significantly higher levels of students were no less out to peers and school self‑esteem75 than students who were out to fewer staff than students in suburban or urban areas, peers76, and this pattern was true for suburban and as indicated in Figure 19.71 More than nine in ten urban students as well.77 Similarly, as shown in students across locales were out to at least one peer, Figure 22, rural LGBT students who were more out and two‑thirds were out to at least one staff member. in school also reported significantly lower levels of

18 Strengths and Silences Strengths and Silences 19 depression78 than other students.79 Nevertheless, Students were also asked how comfortable they were it is important to note that rural students reported speaking with school staff about LGBT issues. Rural poorer well‑being than students from other LGBT students were significantly less comfortable locales — specifically, rural students had lower talking to every type of school personnel than either levels of self‑esteem and higher levels of depression urban or suburban students.84 As shown in Figure than did suburban and urban students, regardless of 25, rural LGBT students were most comfortable how out they were.80 talking with school counselors or teachers, yet even for these two types of staff, only half of respondents For LGBT students in general, being out at school reported feeling comfortable (52% and 48%, was also associated with feeling more connected respectively). Rural LGBT students were least to school, and this pattern was also observed for comfortable talking about these issues with a school rural students. As illustrated in Figure 23, rural safety or resource officer (22%) or a gym teacher or LGBT students who were out to more peers at athletic coach (19%).85 school about their sexual orientation and/or gender identity reported significantly higher levels of school As shown in Figure 26, students were also asked belonging than students who were out to fewer how frequently they had raised LGBT issues with peers.81 The association between outness and school staff. Whereas a little less than half (48%) of belonging was similar across locales, although rural rural students said they were comfortable discussing youth continued to report the lowest levels of school LGBT issues with teachers, a slightly higher belonging regardless of how out they were to peers.82 percentage (57%) said they had actually discussed Together, these findings show that rural students LGBT issues with a teacher. In addition, nearly a were out to peers, school staff, and parents at levels third (31%) had talked with counselors about LGBT similar to suburban and urban students. Being issues in the past year. It is interesting to note that, out was associated with increased victimization with the exception of teachers, students were more for all locales, though even more strongly so for likely to say they would be comfortable talking with rural students. Although rural students appeared the different types of staff than actually had talked to experience a similar increase in well‑being and with them. school belonging when they were out as compared to suburban and urban students, they still reported lower well‑being and school belonging overall than their peers in other locales.

Talking about LGBT Issues in School

Another indicator of school engagement is the degree to which students feel comfortable discussing LGBT issues with school staff and raising LGBT issues in class. When students feel safer in the school environment and feel like a respected part of their school community, they may feel more comfortable raising LGBT issues in class and other school environments. As depicted in Figure 24, about half (55%) of rural LGBT students said they were comfortable raising LGBT issues in class, which is comparable to the 56% of suburban and 57% of urban students who said they were comfortable raising these issues in class.83

20 Strengths and Silences In comparing the frequency of discussing LGBT who did not talk with these staff members about issues across locales, rural students were largely LGBT issues.89 Only for safety/resource officers was similar to urban and suburban students. The only discussing LGBT issues not associated with greater significant differences were that rural students school belonging.90 In that these personnel are often reported slightly lower rates of discussing LGBT used to maintain school safety, this observation issues with teachers86, counselors87, and school might be an indicator of LGBT youth being treated resource/safety officers88 than suburban and urban as perpetrators, or perhaps simply that they speak students. This pattern is surprising, considering that with these people only when they have been in rural students were consistently less comfortable trouble; such instances would understandably not talking with each type of school staff than urban be associated with the increases in school belonging and suburban students. As we found with being out observed for talking with other types of school in school, these findings suggest that rural LGBT personnel. students were relatively open about LGBT issues in the school environment in spite of decreased Together, these findings indicate interesting patterns personal comfort or increased risk for harm. of school engagement for rural LGBT students. Although rural LGBT students reported high levels Being able to speak with school personnel about of victimization, half were comfortable raising LGBT LGBT issues may help LGBT students feel more issues in class, and more than half had discussed connected to their school community. Thus, we LGBT issues with a teacher at least once during the might expect students who talk to staff about LGBT past year. In addition, rural LGBT youth were just issues to feel greater belonging to their school as out in school as suburban and urban students. community. As shown in Figure 27, students who Thus, despite considerable risk involved in being talked with teachers, principals, vice principals, LGBT in rural schools, rural LGBT students reported counselors, nurses, librarians, and gym teachers behaviors that in many ways were similar to their at least once in the past year reported a much suburban and urban counterparts. greater sense of school belonging than students

Strengths and Silences 21 22 Strengths and Silences 11% of rural LGBT students reported having an Resources & Supports LGBT‑inclusive curriculum (i.e., having been taught positive things about LGBT people, history, or events Across locales, LGBT students often encounter in their classes), significantly less than the 18% hostile school settings, and such environments of suburban and 20% of urban students.92 Their can have negative consequences for psychological textbooks and assigned readings were also less likely well‑being and educational success. LGBT‑related to include information about LGBT persons, history, resources can help counter these negative or events (13%, compared to 20% of suburban influences, as well as positively affect the school and 20% of urban students).93 Rural students also climate and enhance the learning environment for reported less access to LGBT‑related content on the LGBT students. Especially given that students in Internet using school computers: 39% of rural LGBT rural areas report more negative school experiences students whose school computers had Internet than those in other locales, it is important to access said that they could access LGBT‑related understand the availability of LGBT‑related resources websites, compared to 44% of suburban students in rural areas of the country. Students in the 2011 and 44% of urban students.94 In addition, only National School Climate Survey were asked about 44% of rural students said they had access to the prevalence of LGBT‑related resources, such LGBT‑related resources in the school library, though as supportive student clubs, curricular resources, their availability was not different from suburban and school policies for addressing harassment and and urban students.95 assault.91 They were also asked about support from peers and school personnel. In this section, we Participants in this survey were also asked about the discuss the availability of resources for rural LGBT sex education provided at their school and whether students and any differences between rural students it used an abstinence‑only approach — that is, that and their urban and suburban peers. students should not have sex until marriage, which is a current impossibility for most LGBT adults in Curricular Resources the U.S. Existing research demonstrates that many abstinence‑only curricula provide misleading and Rural LGBT students have limited access to medically inaccurate information about sexuality LGBT‑related curricular resources, and their access and sexual health, and also that they commonly is consistently lower than students in suburban ignore the needs of LGBT youth, who may not and urban areas. As shown in Figure 28, only receive accurate information about HIV prevention

Strengths and Silences 23 and relevant information on sexual health matters.96 As with curricular resources, GSAs were much Overall, rural (and urban) students were only slightly less prevalent in rural areas than in suburban or less likely than suburban students to have been urban areas: 27% of rural students reported having taught any curriculum about sexual health (84% of a GSA at school, compared to 55% of suburban rural students, vs. 87% of suburban students and students and 53% of urban students.100 However, 83% of urban students).97 However, rural LGBT when there was a GSA at school, rural students students were more likely to have been taught to were more likely to attend than urban and suburban practice abstinence‑only in sex education (see students. As shown in Figure 30, 58% of rural Figure 29). One‑third (32%) of rural LGBT students LGBT students who attended school with a GSA said they were taught an abstinence‑only sexual reported that they attended frequently or often, health curriculum, which was greater than the 28% compared to 49% of urban and 53% of suburban of suburban students and 25% of urban students school students.101 This difference suggests that who were taught abstinence‑only.98 GSAs may serve as a particularly important resource for rural LGBT students, especially when compared Supportive Student Clubs to urban LGBT students. Given that rural students also experienced higher levels of victimization than School clubs supportive of LGBT students and their peers in other locales, it may be that students topics, such as Gay‑Straight Alliances (GSAs), serve in rural areas were likelier to seek out their GSA as another critical resource in creating safe spaces as a safe space — one where they could receive for LGBT students. The 2011 National School support as well as resources to cope with or address Climate Survey, for instance, reported that students victimization.102 in schools with a GSA heard fewer homophobic remarks and were less likely to report feeling unsafe Community programs or groups for LGBT youth or having been victimized because of their sexual that take place outside of the school environment orientation or gender expression than students may serve a similar purpose as school‑based GSAs. who attended schools without GSAs. Furthermore, Unfortunately, rural LGBT students were also much they were more likely to report that teachers and less likely than suburban or urban students to live other staff members intervened when they heard in communities with a program or group for LGBT homophobic remarks.99 youth (30% vs. 44% and 51%, respectively).103 In

24 Strengths and Silences fact, students who attended schools without a GSA were also much less likely to live in a community with a community group or program for LGBT youth (see Figure 31).104 For instance, among rural students, only 21% of students who attended schools without a GSA had a community group or program available as an alternative resource. These findings suggest a shortage of school‑ and community‑based clubs for LGBT students in rural areas. It also suggests, perhaps, that some rural areas may be in even greater need of resources for LGBT youth than others.

Supportive School Personnel

School personnel serve as another important resource for LGBT youth. Having supportive teachers and other school staff has been shown to be positively related to the academic experiences of students in general.105 Because LGBT students, particularly in rural areas, may feel unsafe and experience victimization at school, being able to solicit help from a supportive adult at school may be crucial to creating safer leaning environments. Fortunately, as shown in Figure 32, the vast majority of rural LGBT students (94%) in this survey reported knowing at least one school staff person supportive of LGBT students. However, on average, rural students knew fewer supportive staff members than students in other areas of the country. For instance, only 41% of rural students knew many (6 or more) teachers or staff members supportive of LGBT students, compared to 60% of suburban and 61% of urban students) (see Figure 32).106

School administrators also play an important role in creating a safe school environment because they help establish school policies on harassment and bullying, are in a position to provide training and support to teachers and other staff members regarding LGBT issues, and also help set the tone of the overall school environment. As shown in Figure 33, 25% of rural LGBT students reported having an administration supportive of LGBT students, lower than the 35% of students in suburban schools and 36% in urban schools who said that their administration was supportive of LGBT students.107

Strengths and Silences 25 Supportive Peers School Harassment & Assault Policies

Supportive peers also have the ability to make An additional resource that can substantially the learning environment more positive for LGBT contribute to safer schools for LGBT students is students.108 Rural LGBT students, however, may face the implementation of school anti‑bullying policies significant resistance from peers. As depicted in that specifically enumerate protection based on Figure 33, only 28% of rural students reported that sexual orientation and gender identity/expression. As other students in their schools were accepting of shown in the 2011 National School Climate Survey, LGBT students, indicating that peers in rural areas policies that include these protections were found were significantly less accepting than in suburban to be associated with a lower frequency of hearing (33% accepting) or urban (46% accepting) areas. 109 homophobic remarks in school and negative remarks about gender expression. They were also associated Students in rural areas may also face difficulty in with lower rates of victimization and more effective finding peers who share similar experiences.110 Rural intervention when negative remarks are made.112 students were significantly less likely to have access to a GSA or LGBT community group, for instance Unfortunately, few LGBT students in rural areas (see Figure 28). Although most students regardless reported that their schools had comprehensive of locale knew at least one other LGBT peer, rural harassment and bullying policies that specifically (and suburban) LGBT students knew substantially enumerated sexual orientation and gender identity/ fewer LGBT peers than LGBT students in other expression. As shown in Figure 35, only 18% of areas of the country, even after accounting for the rural students attended schools with policies that smaller sizes of many rural schools. For instance, as enumerated sexual orientation or gender expression, shown in Figure 34, only 52% of students in rural including only 5% who said that their schools areas knew many (six or more) other LGBT students, enumerated both (i.e., had comprehensive policies). compared to 63% of suburban students and 66% of Although comprehensive policies were uncommon urban students.111 regardless of locale, rural students were less likely

26 Strengths and Silences The Benefits of LGBT‑Related Resources and Supports

LGBT‑related resources — student clubs, supportive teachers and staff members, inclusive curricular resources, and comprehensive anti‑bullying and harassment policies — serve as important resources because they have the power to change school cultures and practices. In this section, we discuss the utility of LGBT‑related resources for rural LGBT students. Although these resources have been found to be associated with greater access to education, educational achievement, educational aspirations, and overall school climate for the national LGBT youth population115, it is important to examine their impact in rural environments specifically in order to ensure that safe school efforts are as effective as possible. Accordingly, we examine the relationship between school supports and overall school climate, as well as the relationship between supports and than urban or suburban students to attend schools well‑being and school belonging for rural LGBT with comprehensive policies (5% of rural vs. 8% students. of suburban and 10% of urban students).113 They were also more likely than urban students to have no School Climate policy at all.114 Each of the four LGBT‑related school supports Given that comprehensive policies have been examined here — supportive clubs, supportive staff associated with better school environments, it is members, inclusive curricula, and comprehensive alarming that so few schools — whether rural, policies — was associated with lower levels of suburban, or urban — specifically protect students victimization due to sexual orientation. For instance, on the basis of sexual orientation and gender rural LGBT students who did not have a GSA at their identity/expression. Thus, strategies to improve the school reported a weighted victimization score of school experience for LGBT students must include 5.52, higher than the 3.85 for rural students who efforts to provide them with greater institutional did have a GSA at their school (see Figure 36).116 protections. In addition, rural LGBT students who had many supportive educators experienced lower victimization than students with no supportive educators; rural students with an inclusive curriculum had lower victimization than students without an inclusive curriculum; and those with a comprehensive policy (i.e., which enumerated both sexual orientation and gender identity/expression) experienced lower victimization than students with no anti‑bullying/ harassment policy.117 In a similar fashion, LGBT‑related supports were associated with lower levels of victimization based on gender expression.118

Strengths and Silences 27 Psychological Well‑Being and Connection to School

LGBT‑related resources have also been shown to be related to better psychological well‑being and attachment to school. Findings show that the four LGBT‑related school supports examined here were also associated with better psychological outcomes specifically for rural LGBT students. For instance, rural LGBT students who had many supportive educators reported an average school belonging score of 2.70, higher than the 2.00 observed for students who did not have any supportive educators at their school (see Figure 37).119 In addition, rural LGBT students who had a GSA at their school reported higher belonging than those who did not; rural students with an inclusive curriculum had lower victimization than students without an inclusive curriculum; and those with a comprehensive policy experienced less victimization than students with no anti‑bullying/harassment policy.120 These resources were also found to be associated with lower depression121 and higher self‑esteem among rural LGBT students.122

28 Strengths and Silences Conclusion & Recommendations

race/ethnicity, as most national youth surveys restrict Limitations students to selecting only one racial category, and do not provide a multiracial response option.125 In The methods used for this survey resulted in a contrast, we allow for students in our survey to select nationally representative sample of LGBT youth. multiple options for their race/ethnicity, and code However, it is important to note that the sample is students who selected two or more racial categories representative only of youth who identify as lesbian, as being multiracial.126 gay, bisexual, or transgender, have some connection to the LGBT community (either through their local It is also important to note that our survey only youth organization or through the Internet), and/ reflects the experiences of LGBT students who were or have a Facebook page. Thus, we may not have in school during the 2010–2011 school year. Thus, reached LGBT students who were not connected to findings from this survey may not necessarily reflect LGBT community organizations in some way or who the experiences of LGBT youth who have dropped had limited access to computers or the Internet. out of school, whose experiences with a hostile We also cannot make determinations from our school climate or access to supportive resources may data about the experiences of youth who might be differ from those students who remained in school. engaging in same-sex sexual activity or experiencing same-sex attractions but who do not identify themselves as lesbian, gay, bisexual, or something else other than heterosexual (e.g., queer). Discussion

Just as there are no known population parameters School settings may often pose significant dangers of LGBT youth overall, we cannot be certain as for LGBT students across the country. Findings to the representativeness of our sample of rural in this report suggest that the experiences of LGBT youth relative to the larger population of victimization may be more frequent, and the number rural LGBT youth. Our national sample was slightly of resources fewer, in rural and small town areas more suburban and slightly less rural than the than in other areas of the country. Experiences of population of secondary school students in the victimization and other outcomes were commonly 123 United States. It is possible that our methods had worse for rural students than for suburban and less reach in rural areas, and/or that they resulted in urban students. This stands in contrast to common a rural LGBT youth sample that is more connected depictions of urban schools as the least safe for than the overall rural LGBT youth population. It is students in general. Almost all of the rural LGBT also possible that youth in rural areas who might students in this survey reported feeling unsafe at eventually identify as LGBT as adults do not identify school. Compared to their suburban and urban as such as youth. peers, rural LGBT students experienced more derogatory comments as well as more direct The percentage of youth of color was lower than the harassment in school on the basis of their sexual general population of secondary school students, orientation and/or gender expression. They were also which may be another possible limitation to the more frequently the target of relational bullying, survey. However, our participant outreach methods cyberbullying, and intentional property damage. have resulted in increased representation of youth As we find for all LGBT students, rural students of color over the years, and the characteristics of the also reported that school staff members most often rural LGBT sample were similar to those found in the failed to effectively address their experiences of rural school population in general: more White than victimization. in other areas, more likely to attend public schools and smaller schools, and more likely to live in the Given that attitudes about LGBT people in general 124 South and Midwest. Any discrepancies may also may be less tolerant in rural and small town areas, it have resulted from different methods for measuring is perhaps surprising that rural LGBT students were

Strengths and Silences 29 no less likely to be out to school staff and peers than In addition, GSAs may serve particular importance suburban and urban LGBT students. However, our in rural areas. Compared to students in urban or report reveals that the consequences they face for suburban areas, students in rural areas were more being open about being LGBT are more severe than likely to attend GSAs when their schools had them. those experienced by suburban and urban students, When they are present, GSAs may help rural LGBT however. Although being more out at school was students build a support network of other LGBT related to higher levels of victimization for LGBT students and supportive peers. Nevertheless, students in general, the relationship was even findings from this report highlight the heightened stronger for rural LGBT students. potential for isolation among LGBT students in rural areas. LGBT youth in rural and small town Even though this report depicts substantially more areas were less likely to have access to an LGBT negative school experiences for rural LGBT students community group, less likely to have a GSA in their compared to suburban and urban students, it is not school, and knew fewer LGBT peers at school, than meant to suggest that rural and small town areas of LGBT students in other locales. Thus, these findings the country are universally negative places for LGBT also indicate continued need for LGBT-related persons. Indeed, a small but enlightening body of resources in rural and small town areas. scholarship has examined how LGBT people in rural areas of the country resist characterizations of their experiences as inferior to those in urban locales and create strong and meaningful communities.127 Future Directions Nonetheless, it does suggest that although rural LGBT youth are at times resilient, they continue to for Research face stigma, greater victimization, and lower access to resources. Although most LGBT students overall This report fills an important gap in our knowledge reported a lack of positive, LGBT-related resources of the experiences of rural and small town LGBT in their school, this problem was particularly youth. Although we provide a broad perspective of pronounced for students in rural and small town the experiences of LGBT students living in rural areas. Rural LGBT students were far less likely to areas, more can be learned about differences within have a GSA in their school, to say that their school the rural LGBT student population. Future research curricula were inclusive of LGBT persons and events, should examine how racial/ethnic identity intersects and to report having a comprehensive anti-bullying with locale and other characteristics in influencing policy at school. Furthermore, they were less likely the experiences of LGBT youth. Similarly, it would to report that the administration at their school be important to understand the unique experience of was supportive of LGBT students. Even though transgender students in rural areas. Research should the vast majority of rural LGBT students reported consider the role that additional factors play in knowing at least one school staff person supportive communities and in the school experiences of LGBT of LGBT students, rural students still reported fewer students, such as religion, socioeconomic status, supportive educators, on average, than students in and local traditions and culture. National population- rural and urban areas. based surveys of youth, such as the Youth Risk Behavior Survey, need to include questions about It is important to note that rural LGBT students sexual orientation and gender identity to allow for benefit from LGBT-related school and community more in-depth examination of the experiences of resources, particularly school personnel and GSAs. rural student experiences. Despite feeling uncomfortable talking with school staff about LGBT issues, rural LGBT students Given that rural students reported substantially nonetheless discussed these issues with staff as higher levels of victimization than students in frequently or nearly as frequently as students from other areas of the country and substantially fewer other areas of the country. Moreover, discussing LGBT‑related resources than other youth, it is these issues was associated with greater school important for future research to examine whether belonging for nearly every type of school personnel. different interventions are more or less effective These findings suggest that school personnel serve in different types of communities, particularly as important resources in rural schools, even if across different locales. In addition, more students are less comfortable approaching them knowledge is needed about possible differences in about LGBT issues. implementation across locales — certain types of

30 Strengths and Silences interventions may be less difficult or controversial It is important to acknowledge the challenges to implement in rural and small town settings. posed by limited resources and the low population Furthermore, more formative research is needed that densities of rural areas. These factors may examines new strategies for creating safer school necessitate innovative approaches to making climates for LGBT students in rural schools. Given schools safer for rural LGBT students. For instance, that LGBT rural students reported having fewer increasing access to LGBT‑related resources through LGBT-related supports, it is important for future computers may help students access resources and research to examine both the resilience of students information that might not otherwise be available experiencing hostile school climates as well as or offered in the classroom. Rural educators might strategies used by rural students to seek support, also advocate for new or additional technologies that perhaps especially when they encounter resistance would provide access to online communities and in the school and community environments. supports for more geographically (and otherwise) isolated youth.

LGBT students were most comfortable talking Recommendations to counselors and teachers about LGBT issues. However, the prevalence of supportive educators These findings demonstrate a clear need for was lower in rural areas; thus, there is a need for safer and more inclusive learning environments greater emphasis on professional development for for LGBT students in rural and small town areas. school professionals in rural areas. In addition to Anti-LGBT bullying is a problem in many areas of professional development, it may be even more the country, but this report suggests that efforts to important for supportive educators to identify address anti-LGBT bullying may require particular themselves more intentionally as allies, through attention in rural areas. Educators, policymakers, public support for LGBT students, such as use of and supporters of safe school initiatives can use the GLSEN’s Safe Space stickers and posters, inclusion information in this report to better understand the of LGBT content in class materials, and support specific experiences of rural LGBT students and take for student events such as Ally Week or the Day of appropriate steps to make rural schools safer and Silence. Given that rural areas appear most likely more inclusive for LGBT students. to lack GSAs and community-based LGBT youth groups, it may be even more important for supportive The 2011 National School Climate Survey educators in rural schools to consider sponsoring a found that school- and community-based GSA or a similar club supportive of LGBT student resources — GSAs, supportive staff, comprehensive issues. policies, inclusive curricula, and LGBT youth groups — are associated with safer schools for Together, these recommendations will help make LGBT students.128 Unfortunately, rural schools are schools safer for all students in school, regardless much less likely to have these kinds of resources of sexual orientation, gender identity, gender and supports than suburban and urban areas. expression, or locale. Given that rural and small town areas may be more politically and socially conservative, they may also be somewhat resistant to implementation of these measures in the future. Therefore, pursuing LGBT-related resources and supports may require alternative strategies in rural areas. Education leaders and safe school advocates may find it useful to partner with other organizations to create a broader movement to advocate for safe schools, which include LGBT-specific policies and practices. Members of rural and small town LGBT communities may also find it valuable to establish community groups and programming for LGBT youth, as they seem to be particularly absent from rural and small town areas.

Strengths and Silences 31

Notes & References

1 Dinkes, R., Kemp, J., & Baum, K. (2009). Indicators 7 Davis, J., & Smith, T. (2004). General Social Surveys, of school crime and safety: 2008 (NCES 2009-022/ 1972-2004: Cumulative codebook. Chicago: National NCJ 226343). Washington, DC: U.S. Department of Opinion Research Center. Education, Institute of Education Sciences, National Center for Education Statistics, and U.S. Department of Dillon, M., & Savage, S. (2006). Values and religion in Justice, Office of Justice Programs, Bureau of Justice rural America: Attitudes toward abortion and same-sex Statistics.. relations. In The Carsey Institute reports on rural America (pp. 1-10). Durham, NH: Carsey Institute, University of 2 Kosciw, J. G., Greytak, E. A., & Diaz, E. M. (2009). New Hamshire. Who, what, where, when, and why: Demographic and ecological factors contributing to hostile school climate Glenna, L. (2003). Religion. In D. L. Brown & L. E. for lesbian, gay, bisexual, and transgender youth. Journal Swanson (Eds.), Challenges for rural America in the of Youth and Adolescence, 38, 976–988. twenty-first century(pp. 262-272). University Park, PA: The Pennsylvania State University Press. 3 Pew Research Center for the People and the Press. (2003). Republicans unified, Democrats split on gay Rogers, E. M., Burdge, R. J., Korsching, P., & marriage: Religious beliefs underpin opposition to Donnermeyer, J. (1988). Social change in rural societies: homosexuality. Washington, D.C.: The Pew Research An introduction to rural sociology (3rd ed.). Englewood Center for the People and the Press. Cliffs, NJ: Prentice Hall.

4 Dillon, M., & Henly, M. (2008, Fall). Religion, politics, Smith, J. D. (1997). Working with larger systems: Rural and the environment in rural America. University of New lesbians and gays. In J. D. Smith & R. J. Mancoske Hampshire, Carsey Institute. (Eds.), Rural gays and lesbians: Building on the strength of communities (pp. 13-22). New York, NY: The Lichter, D. T., & Crowley, M. L. (2002). Poverty in Harrington Park Press. America: Beyond welfare reform. Population Bulletin, 57(2), 1-36. Swank, E. D., Frost, D. M., & Fahs, B. (2012). Rural location and exposure to minority stress among sexual Lugo, L., Stencel, S., Green, J., Smith, G., Cox, D., minorities in the United States. Psychology & Sexuality, Miller, T., et al. (2008). U.S. religious landscape survey. 3(3), 226-243. Pew Forum on Religion and Public Life and The Pew Research Center. Washington, D.C.: The Pew Forum. 8 GLSEN and Harris Interactive (2012). Playgrounds & prejudice: Elementary school climate in the United 5 Pew Research Center for the People and the Press. States. New York: Gay, Lesbian & Straight Education (2012). Changing views of gay marriage: A deeper Network. analysis. Retrieved 23. October 2012 from http://www. people-press.org/2012/05/23/changing-views-of-gay- 9 Fetner, T., & Kush, K. (2008). Gay-straight alliances in marriage-a-deeper-analysis/ high schools: Social predictors of early adoption. Youth & Society, 40(1), 114-130. 6 Klawitter, M., & Hammer, B. (1999). Spatial and temporal diffusion of local antidiscrimination policies 10 Stapel, C. J. (2008). No longer alone: A resource manual for sexual orientation. In E. D. B. Riggle & B. L. Tadlock for rural sexual minority youth and the adults who serve (Eds.), Gays and lesbians in the democratic process: them. Cambridge, MA: Harvard Gay and Lesbian Causus. Public policy, public opinion, and political representation Retrieved from www.ruralgayyouth.com. (pp. 22-38). New York: Columbia University Press. 11 Smith, J. D. (1997). Working with larger systems: Rural lesbians and gays. In J. D. Smith & R. J. Mancoske (Eds.), Rural gays and lesbians: Building on the strength of communities (pp. 13-22). New York, NY: The Harrington Park Press.

Yarbrough, D. G. (2004). Gay adolescents in rural areas. Journal of Human Behavior in the Social Environment, 8(2-3), 129-144.

Strengths and Silences 33 12 Gray, M. L. (2010). From websites to Wal-Mart: Youth, 18 To test differences across locale in hearing biased identity work, and the queering of boundary publics in languages, several analyses (ANOVAs) of variance were Small Town, USA. In C. Pullen & M. Cooper (Eds.), LGBT conducted, with frequency of heading the biased remark identity and online new media (pp. 299-298). New York: as the dependent variable and locale as the independent Routledge. variable. The main effects were significant for hearing “gay” in a negative way: F(2, 8144)=56.32, p<0.001, Gray, M. L. (2009). Out in the country: Youth, media, effect size =.014; hearing other homophobic remarks; and queer visibility in rural America. New York: NYU F(2, 8137)=85.16, p<0.001, effect size =.020; hearing Press. remarks about someone not acting masculine enough: F(2, 8127)=22.37, p<0.001, effect size =.005; hearing 13 Davis, J., & Smith, T. (2004). General Social Surveys, remarks about someone not acting feminine enough: 1972-2004: Cumulative codebook. Chicago: National F(2, 8127)=16.13, p<0.001, effect size =.004; hearing Opinion Research Center. racist remarks: F(2, 8135)=8.29, p<0.001, effect size =.002; and hearing sexist remarks: F(2, 8124)=4.56, Dillon, M., & Savage, S. (2006). Values and religion in p<0.01, effect size =.001. rural America: Attitudes toward abortion and same-sex relations. In The Carsey Institute reports on rural America 19 To test differences across locale, an analysis of variance (pp. 1-10). Durham, NH: Carsey Institute, University of (ANOVA) was conducted with frequency of hearing the New Hamshire. phrase “no homo” as the dependent variable and locale as the independent variable. The main effect of locale 14 Locale information was captured by asking students for was significant: F(2, 8143)=3.52,p <0.05, effect size their school district and/or zip code. These were then =.001. matched with NCES school district-level information. 20 Brown, J. R. (2011). No homo. Journal of 15 In this study, “rural” refers to rural and small town Homosexuality, 58(3), 299-314. students. Statistical comparison tests suggested that students from these areas could be treated similarly in 21 Based on paired sample comparison tests: students said statistical analysis; in addition, the aggregation into a teachers were more likely to intervene when they heard single “rural” category generates a sample size adequate racist remarks than homophobic remarks (t=31.553, for subsample comparisons. Categories “suburban” and p<0.001), sexist remarks (t=15.565, p<0.001), “urban” were aggregated in similar ways and for similar and negative remarks regarding gender expression reasons. (t=29.024, p<0.001).

16 To test demographic differences across locale, several 22 To test differences in rates of intervention in biased analyses of variance (ANOVAs) were conducted. To remarks across locales, a multivariate analysis test differences in race, an analysis of variance was of variance (MANOVA) was conducted with staff conducted with race as the dependent variable and intervention most of the time or always as the dependent locale as the independent variable. The main effect of variable, and locale as the independent variable. The locale was significant: F(2, 7845)=94.657,p <0.001, multivariate effect was significant: Pillai’s trace=.004, effect size=.024. To test differences in school type, F(8, 7924)=1.945, p<.05. The univariate effect of an analysis of variance was conducted with attending locale in intervening in homophobic remarks was a public school as the dependent variable and locale significant: F(2, 3964)=4.881,p <0.01. Post-hoc as the independent variable. The main effect of locale Bonferroni tests indicated that suburban staff members was significant: F(2, 7845)=56.863,p <0.001, effect were more likely to intervene than rural staff members. size=.014. To test differences in region, an analysis of Urban staff members were not different from either variance was conducted with south, midwest, west, and group. The univariate effect of locale in intervening in northeast as the dependent variables, and locale as the racist remarks was not significant: F(2, 3964)=.326, independent variable. The main effect of locale was p>0.10. The univariate effect of locale in intervening in significant at thep <0.001 for each region, with effect sexist remarks was not significant: F(2, 3964)=1.472, sizes ranging from .010 to .039. To test differences p>0.10. The univariate effect of locale in intervening in school size, an analysis of variance was conducted in negative remarks about gender expression was not with school size as the dependent variable and locale significant: F(2, 3964)=0.099,p >0.10. as the independent variable. The main effect of locale was significant: F(2, 7845)=351.009,p <0.001, effect 23 Based on paired sample comparison tests: students size=.082. Rural, suburban, and urban samples did not were significantly less likely to intervene than teachers differ demographically in other respects. in homophobic remarks (t=20.026, p<0.001), racist remarks (t=43.345, p<0.001), sexist remarks 17 Based on paired sample comparison tests: students were (t=25.832, p<0.001), and negative remarks regarding significantly more likely to hear biased language from gender expression (t=9.676, p<0.001). students than from teachers for homophobic remarks (t=81.863, p<0.001), racist remarks (t=58.032, 24 Based on paired sample comparison tests: students were p<0.001), sexist remarks (t=62.689, p<0.001), more likely to intervene when they heard racist remarks and negative remarks regarding gender expression than homophobic remarks (t=13.637, p<0.001), and (t=52.635, p<0.001). negative remarks regarding gender expression (t=14.033, p<0.001). They were also more likely to intervene when they heard sexist remarks than in homophobic remarks (t=13.219, p<0.001) and negative remarks regarding gender expression (t=14.195, p<0.001).

34 Strengths and Silences 25 To test differences in rates of intervention in biased 31 To test differences in feeling unsafe due to gender across remarks across locales, a multivariate analysis of locale, an analysis of variance (ANOVA) was conducted variance (MANOVA) was conducted with student with the proportion of students feeling unsafe due to intervention most of the time or always as the dependent their gender as the dependent variable, locale as the variable, and locale as the independent variable. The independent variable, and sexual orientation and gender multivariate effect was marginally significant: Pillai’s identity as covariates. The main effect of locale was not trace=.002, F(8, 14060)=1.788, p<.10. The univariate significant: F(2, 7993)=.199,p >.10. To test differences effect of locale in intervening in homophobic remarks in feeling unsafe due to race across locale, an analysis of was only marginally significant: F(2, 7032)=2.328, covariance (ANCOVA) was conducted with the proportion p<0.10; the differences were not all that meaningful of students feeling unsafe due to their race as the however, as only 6% of urban students were reported to dependent variable, locale as the independent variable, intervene most or all of the time to homophobic remarks, and race as a covariate. The main effect of locale compared to 5% of rural students and 5% of suburban was significant: F(2, 7974)=7.3162,p <.001, effect students. The univariate effect of locale in intervening in size=.002. To test differences in feeling unsafe due to a racist remarks was not significant: F(2, 7032)=2.277, disability across locale, an analysis of variance (ANOVA) p>0.10. The univariate effect of locale in intervening in was conducted with the proportion of students feeling sexist remarks was not significant: F(2, 7032)=2.250, unsafe due to their disability as the dependent variable, p>0.10. The univariate effect of locale in intervening and locale as the independent variable. The main effect in negative remarks about gender expression was not of locale was not significant: F(2, 8025)=.149,p >.10. significant: F(2, 7032)=0.992,p >0.10. 32 Based on paired sample comparison tests: rural students 26 A chi-square test was conducted to compare the were significantly more likely to avoid bathrooms and percentages of students who reported feeling unsafe locker rooms than the cafeteria, school athletic fields, at all by locale: χ2=77.041, df=2, p<.001, Ф=.098. school buses, school grounds, and other places. For Adjusted p-values indicated that rural students were every comparison, p<0.001. more likely to report having feeling unsafe than urban and suburban students. 33 Based on paired sample comparison tests conducted separately for suburban and urban students: students 27 To test differences across locale in feeling unsafe due were significantly more likely to avoid bathrooms and to sexual orientation, an analysis of covariance (ANOVA) locker rooms than the cafeteria, school athletic fields, was conducted with the proportion of students feeling school buses, school grounds, and other places. For unsafe due to their sexual orientation as the dependent every comparison, p<0.001. variable, locale as the independent variable, and sexual orientation as a covariate. The main effect of locale 34 To test differences in avoiding spaces at school by was significant: F(2, 8013)=44.793,p <0.001, effect locale, a multivariate analysis of variance (MANOVA) size=.011. was conducted with avoiding spaces as the dependent variable and locale as the independent variable. The 28 To test differences across locale in feeling unsafe due to multivariate effect was significant: Pillai’s trace=.014, gender expression, an analysis of covariance (ANCOVA) F(18, 15816)=6.297, p<.001. The univariate effect was conducted with the proportion of students feeling of locale was significant in avoiding bathrooms: F(2, unsafe due to their gender expression as the dependent 7915)=22.446, p<0.001, effect size =.006; locker variable, locale as the independent variable, and gender rooms: F(2, 7915)=26.854, p<0.001, effect size =.007; identity and sexual orientation as covariates. The main hallways: F(2, 7915)=13.468, p<0.001, effect size effect of locale was significant: F(2, 7993)=18.611, =.003; school grounds: F(2, 7915)=8.470, p<0.001, p<0.001, effect size=.005. effect size =.002; school athletic fields or facilities: F(2, 7915)=9.384, p<0.001, effect size=.002; and 29 Based on effect sizes for locale in the preceding physical education or gym class: F(2, 7915)=13.588, analyses. effect size=.003. The univariate effect for locale was not significant for avoiding other places: F(2, 7915)=0.810, 30 To test differences across locale in feeling unsafe due p>0.10; or for avoiding the cafeteria: F(2, 7915)=0.283, to religion, an analysis of covariance (ANCOVA) was p>0.10. conducted with the proportion of students feeling unsafe due to their religion as the dependent variable, 35 Based on effect sizes reported above. locale as the independent variable, and religion as a covariate. The main effect of locale was significant: F(2, 7951)=39.160, p<0.001, effect size=.010.

Strengths and Silences 35 36 Davis, J., & Smith, T. (2004). General Social Surveys, 41 Based on paired sample comparison tests of the 1972-2004: Cumulative codebook. Chicago: National frequency of verbal harassment due to different Opinion Research Center. characteristics, students were significantly more likely to be verbally harassed based on their gender expression Dillon, M., & Savage, S. (2006). Values and religion in than based on their gender (t=24.444, p<0.001), race rural America: Attitudes toward abortion and same-sex (t=37.285, p<0.001), disability (t=41.092, p<0.001), relations. In The Carsey Institute reports on rural America or religion (t=19.872, p<0.001). (pp. 1-10). Durham, NH: Carsey Institute, University of New Hamshire. 42 Based on paired sample comparison tests of the frequency of physical harassment due to different Lindhorst, T. (1997). Lesbians and gay men in the characteristics, students were significantly more country: Practice implications for rural social workers. likely to be physically harassed based on their sexual In J. D. Smith & R. J. Mancoske (Eds.), Rural gays and orientation than based on their gender (t=25.264, lesbians: Building on the strength of communities (pp. p<0.001), gender expression (t=16.321, p<0.001), race 1–11). New York, NY: The Harrington Park Press. (t=32.255, p<0.001), disability (t=31.984, p<0.001), or religion (t=29.153, p<0.001). Swank, E. D., Frost, D. M., & Fahs, B. (2012). Rural location and exposure to minority stress among sexual 43 Based on paired sample comparison tests of the minorities in the United States. Psychology & Sexuality, frequency of physical harassment due to different 3(3), 226-243. characteristics, students were significantly more likely to be physically harassed based on their gender expression 37 Pew Research Center for the People and the Press. than based on their gender (t=14.237, p<0.001), race (2012). Changing views of gay marriage: A deeper (t=23.058, p<0.001), disability (t=23.116, p<0.001), analysis. Retrieved 23. October 2012 from http://www. or religion (t=18.455, p<0.001). people-press.org/2012/05/23/changing-views-of-gay- marriage-a-deeper-analysis/ 44 Based on paired sample comparison tests of the frequency of physical assault due to different 38 To test differences in feeling unsafe based on sexual characteristics, students were significantly more orientation across locale and region, a two-way analysis likely to be physically assaulted based on their sexual of variance (ANOVA) was conducted with feeling unsafe orientation than based on their gender (t=16.897, as the dependent variable, and locale and region as p<0.001), gender expression (t=11.326, p<0.001), race independent variables. The main effect of locale was (t=20.491, p<0.001), disability (t=20.197, p<0.001), significant: F(2, 7964)=38.625,p <0.001, as was the or religion (t=18.425, p<0.001). main effect for region: F(3, 7964)=19.011, p<0.001. Bonferroni post-hoc tests indicated that students I 45 Based on paired sample comparison tests of the nthe South and Midwest felt less safe than students in frequency of physical assault due to different the Northeast and West, as did students in rural areas characteristics, students were significantly more likely to compared to students in suburban and urban areas. The be physically assaulted based on their gender expression overall interaction effect between region and locale was than based on their gender (t=9.872, p<0.001), race not significant, although individual interaction effects (t=15.266, p<0.001), disability (t=14.847, p<0.001), indicated that differences between rural and urban or religion (t=11.679, p<0.001). areas were larger in the Northeast (p<.05) and Midwest regions(p<.10) than in the West, where the interaction 46 Using a weighted victimization score based on effect was not different from the South. experiences of verbal harassment, physical harassment, and physical assault. 39 To test differences in feeling unsafe based on sexual orienation across region for rural students, an analysis 47 To test differences in victimization based on sexual of variance (ANOVA) was conducted with feeling unsafe orientation across locale, an analysis of covariance as the dependent variable, and region as independent (ANCOVA) was conducted with the mean weighted variable. The main effect for region was significant: victimization score as the dependent variable, locale F(3, 2330)=5.836, p<0.001. Post-hoc Bonferroni as the independent variable, and sexual orientation as tests indicated that levels of feeling unsafe did not a covariate. The main effect of locale was significant: differ between rural students in the South and Midwest, F(2, 8060)=44.959, p<0.001, effect size=.011. In though both of these groups reported higher levels of addition, post-hoc Bonferroni tests indicated that levels feeling unsafe than did rural students in the Northeast or of victimization based on sexual orientation did not differ West. between urban and suburban students.

40 Based on paired sample comparison tests of the 48 To test differences in victimization based on gender frequency of verbal harassment due to different across locale, an analysis of covariance (ANCOVA) was characteristics, students were significantly more conducted with the mean weighted victimization score likely to be verbally harassed based on their sexual as the dependent variable, locale as the independent orientation than based on their gender (t=42.966, variable, and gender and sexual orienation as p<0.001), gender expression (t=23.288, p<0.001), race covariates. The main effect of locale was significant: (t=58.705, p<0.001), disability (t=62.480, p<0.001), F(2, 7961)=6.722, p<0.001, effect size=.002. In or religion (t=39.475, p<0.001). addition, post-hoc Bonferroni tests indicated that levels of victimization based on gender did not differ between urban and suburban students.

36 Strengths and Silences 49 To test differences in victimization based on gender 55 To test differences in reporting incidents of harassment expression across locale, an analysis of covariance and assault across locales, an analysis of variance (ANCOVA) was conducted with the mean weighted (ANOVA) was conducted with rate of reporting as the victimization score as the dependent variable, locale dependent variable and locale as the independent as the independent variable, and gender and sexual variable. The main effect of locale in reporting to school orienation as covariates. The main effect of locale staff was not significant: F(2, 6191)=1.837,p >0.10. was significant: F(2, 7841)=21.251,p <0.001, effect size=.005. In addition, post-hoc Bonferroni tests 56 To test differences in reporting incidents of harassment indicated that levels of victimization based on gender and assault across locales, an analysis of variance expression did not differ between urban and suburban (ANOVA) was conducted with effectiveness of staff students. response as the dependent variable and locale as the independent variable. The main effect of locale 50 To test differences in victimization based on religion in effectiveness of response was significant: F(2, across locale, an analysis of covariance (ANCOVA) was 2459)=8.827, p<0.001. Post-hoc Bonferroni tests conducted with the mean weighted victimization score suggested that rural students rated staff response as less as the dependent variable, locale as the independent effective than suburban students. Suburban and urban variable, and religion as a covariate. The main effect of students did not differ from one another in their ratings locale was significant: F(2, 7828)=26.660,p <0.001, of effectiveness. effect size=.007. In addition, post-hoc Bonferroni tests indicated that levels of victimization based on religion 57 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, did not differ between urban and suburban students. Madelyn J., & Palmer, N. A. (2012). The 2011 National School Climate Survey: The experiences of lesbian, gay, 51 Based on the effect sizes reported in the notes in the bisexual, and transgender youth in our nation’s schools. first part of this section. New York: Gay, Lesbian & Straight Education Network.

52 To test differences in victimization based on disability 58 A chi-square test was conducted to compare the across locale, an analysis of variance (ANOVA) was percentages of students who had missed a day of school conducted with the mean weighted victimization score due to feeling unsafe by locale: χ2=31.134, df=2, as the dependent variable and locale as the independent p<.001, Ф=.062. Adjusted p-values indicated that rural variable. The main effect of locale was not significant: students were more likely to report having missed school F(2, 7949)=.832, p>0.10. than urban and suburban students.

53 To test differences in victimization based on race 59 To test differences in missing school by severity of across locale, an analysis of covariance (ANCOVA) was verbal harassment, an analysis of variance (ANOVA) conducted with the mean weighted victimization score was conducted with missing school at least once in the as the dependent variable, locale as the independent past month as the dependent variable, and severity of variable, and race and sexual orientation as covariates. verbal harassment as the independent variable. The The main effect of locale was significant: F(2, main effect for severity of verbal harssment based on 7882)=5.197, p<0.01, effect size=.006. In addition, sexual orientation was significant: F(1, 2365)=264.696, post-hoc Bonferroni tests indicated that levels of p<.001. The main effect for severity of verbal harssment victimization based on race were higher for urban than based on gender expression was significant: F(1, for suburban and rural students, who did not differ from 2329)=127.314, p<.001. one another. 60 To test differences in GPA for rural LGBT students by 54 To test differences in frequncies of other types of severity of verbal harassment, an analysis of covariance harassment across locales, a multivariate analysis of (ANCOVA) was conducted with the GPA as the variance (MANOVA) was conducted with frequency of dependent variable, severity of verbal harassment as the the different types of harassment as the dependent independent variable, and frequency of school absences variables and locale as the independent variable. The as a covariate. The main effect for severity of verbal multivariate effect was significant: Pillai’s trace=.019, harassment based on sexual orientation was significant F(10, 15964)=15.246, p<.001. The univariate effect at the p<0.01 level for harassment based on sexual of locale in sexual harassment was significant: F(2, orientation, and significant at thep <0.001 level for 7985)=23.992, p<0.001, effect size =.006. The harassment based on gender expression. univariate effect of locale in rumors or lies spread was significant: F(2, 7985)=62.687,p <0.001, effect 61 To test differences in educational aspirations by size =.015. The univariate effect of locale in feeling severity of victimization based on gender expression, excluded or left out was significant: F(2, 7985)=42.459, an analysis of variance (ANOVA) was conducted with p<0.001, effect size =.011. The univariate effect of plans to attend college as the dependent variable and locale in stolen or damaged property was significant: severity of victimization based on gender expression F(2, 7985)=16.699, p<0.001, effect size =.004. as the independent variable. The effect for severity of The univariate effect of locale in cyberbullying was verbal harassment was significant: F(1, 2320)=4.922, significant: F(2, 7985)=38.050,p <0.001, effect p<0.05. The effect for severity of physical harassment size=.009. was significant: F(1, 2330)=22.039,p <0.001. The effect for severity of physical assault was significant: F(1, 2337)=11.987, p<0.001.

Strengths and Silences 37 62 To test differences in educational aspirations by 68 To test differences across locale, an analysis of severity of victimization based on sexual orientation, covariance (ANCOVA) was conducted with sense of an analysis of variance (ANOVA) was conducted with school belonging as the dependent variable; locale as the plans to attend college as the dependent variable and independent variable; and weighted victimization scores severity of victimization based on sexual orientation based on sexual orientation and gender expression, as as the independent variable. The effect for severity of well as frequency of homophobic and gender-expression- verbal harassment was significant: F(1, 2356)=17.627, related biased remarks as covariates. The main effect p<0.001. The effect for severity of physical harassment for locale remained significant: F(2, 7721)=15.620, was significant: F(1, 2350)=16.043,p <0.001. The p<0.001. Post-hoc Bonferroni tests indicated that after effect for severity of physical assault was significant: F(1, accounting for locale differences in victimization and 2342)=17.348, p<0.001. biased remarks, rural students still had lower levels of school belonging than either suburban or urban students. 63 A chi-square test was conducted to compare the There were no differences between suburban and urban percentages of students who reported discriminatory students on levels of school belonging. policies and practices by locale: χ2=23.587, df=2, p<.001, Ф=.054. Adjusted p-values indicated that rural 69 Hershberger, S. L., & D’Augelli, A. R. (1995). The students were more likely to report having discriminatory impact of victimization on the mental health and policies and practices than urban and suburban suicidality of lesbian, gay, and bisexual youths. students. Developmental Psychology, 31(1), 65-74.

64 Goodenow, C., & Grady, K. E. &1993). The relationship Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, of school belonging and friends’ values to academic Madelyn J., & Palmer, N. A. (2012). The 2011 National motivation among urban adolescent students. Journal of School Climate Survey: The experiences of lesbian, gay, Experimental Education, 62(1), 60-71. bisexual, and transgender youth in our nation’s schools. New York: Gay, Lesbian & Straight Education Network. Murdock, T. B., & Bolch, M. B. (2005). Risk and protective factors for poor school adjustment in lesbian, Rosario, M., Hunter, J., Maguen, S., Gwadz, M., & gay, and bisexual (LGB) high school youth: Variable and Smith, R. (2001). The coming-out process and its person-centered analyses. Psychology in the Schools, adaptational and health-related associations among gay, 42(2), 159-172. lesbian, and bisexual youths: Stipulation and exploration of a model. American Journal of Community Psychology, Roeser, R. W., Midgley, C., & Urdan, T. C. (1996). 29(1), 133-160. Perceptions of the school psychological environment and early adolescents’ psychological and behavioral 70 Based on paired sample comparison tests: using functioning in school: The mediating role of goals and a 4-point outness Likert-type scale, students were belonging. Journal of Educational Psychology, 88, 408- significantly more likely to be out to peers than to school 422. staff (t=-43.688, p<0.001). Using a binary measure of being out to at least one person, students were 65 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, siginificantly more likely to be out to students than staff Madelyn J., & Palmer, N. A. (2012). The 2011 National (t=29.752, p<0.001). School Climate Survey: The experiences of lesbian, gay, bisexual, and transgender youth in our nation’s schools. 71 To test differences in outness across locale, a New York: Gay, Lesbian & Straight Education Network. multivariate analysis of variance (MANOVA) was conducted with being out to at least one peer and one 66 A measure for the psychological sense of school staff personas the dependent variables and locale as membership was developed for use with aadolescents the independent variable. The multivariate effect was by Carol Goodenow: Goodenow, C. (1993). The marginally significant and small: Pillai’s trace=.001, psychological sense of school membership among F(4, 16142)=2.071, p<.10. The univariate effect for adolescents: Scale development and educational locale in outness to peers was marginally significant: correlates. Psychology in the Schools, 30(1), 79-90. The F(2, 8071)=2.614, p<0.10. The univariate effect for measure used here is based on the 4-point composite locale in outness to school staff was not significant: F(2, scale constructed from these measures of school 8071)=2.247, p>0.10. belongingnes. 72 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, 67 To test differences across locale, an analysis of variance Madelyn J., & Palmer, N. A. (2012). The 2011 National (ANOVA) was conducted with sense of of school School Climate Survey: The experiences of lesbian, gay, belonging as the dependent variable and locale as the bisexual, and transgender youth in our nation’s schools. independent variable. The main effect for locale was New York: Gay, Lesbian & Straight Education Network. significant: F(2, 8114)=69.364,p <0.001. Post-hoc Bonferroni tests indicated that rural students had lower levels of school belonging than either suburban or urban students. There were no differences between suburban and urban students on their levels of school belonging.

38 Strengths and Silences 73 To test differences in victimization for rural students 79 To test differences in depression by level of outness by level of outness to peers, a multivariate analysis of for rural students, an analysis of variance (ANOVA) was variance (MANOVA) was conducted with victimization conducted with depression as the dependent variable based on sexual orientation and victimization based and outness to peers as the independent variable. The on gender expression as the dependent variables, and main effect for outness to peers was significant: F(1, outness to peers as the independent variable. The 2324)=18.133, p<0.001. Post-hoc Bonferroni tests multivariate effect was significant: Pillai’s trace=.044, indicated that rural students reported higher levels of F(2, 2272)=52.654, p<.001. The univariate effect depression than suburban and urban students, who were for outness in sexual orientation victimization was not different from one another. significant: F(1, 2273)=87.842,p <0.001. The univariate effect for outness in gender expression 80 To test differences in self-esteem across locale and victimization was significant: F(1, 2273)=17.356, outness to peers, an analysis of covariance (ANCOVA) p<0.001. was conducted with self-esteem as the dependent variable, locale and outness to peers as independent 74 To test differences in victimization by locale and outness variables, and interactions between locale and outness. to peers, an analysis of covariance (ANCOVA) was The main effect for locale was significant: F(2, conducted with victimization based on sexual orientation 7953)=3.938, p<0.05. The interaction effect was not as the dependent variable, locale as the independent significant. Post-hoc Bonferroni tests indicated that variable, and interactions between locale and outness. rural students reported lower levels of self-esteem than The main effect for outness in sexual orientation suburban students, and lower levels than urban students victimization was significant: F(1, 8011)=191.172, with marginal statistical significance. Suburban students p<0.001. The interaction effect for outness was were not statistically different from urban students. significant overall at thep <.001 level, and was also significant in individual comparisons between rural and 81 To test differences in school belonging by level of suburban and urban students. outness for rural students, an analysis of variance (ANOVA) was conducted with sense of school belonging 75 Self-esteem was assessed with the Rosenberg Self- as the dependent variable and outness to peers as the Esteem Scale. Rosenberg, M. (1989). Society and the independent variable. The main effect for outness to adolescent self-image (Revised edition). Middletown, CT: peers was significant: F(1, 2363)=26.981,p <0.001. Wesleyan University Press. 82 To test differences in school belonging across locale and 76 To test differences in self-esteem by level of outness outness to peers, an analysis of covariance (ANCOVA) for rural students, an analysis of variance (ANOVA) was was conducted with sense of belonging as the dependent conducted with self-esteem as the dependent variable variable, locale as the independent variable and outness and outness to peers as the independent variable. The to peers as a covariate. The main effect for locale and main effect for outness to peers was significant: F(1, was significant: F(2, 8065=71,232,p <0.001. 2335)=51.648, p<0.001. 83 To test differences in comfort in raising LGBT issues 77 To test differences in self-esteem across locale and in class by locale, an analysis of variance (ANOVA) outness to peers, an analysis of covariance (ANCOVA) was conducted with comfort in raising LGBT issues as was conducted with self-esteem as the dependent the dependent variable and locale as the independent variable, locale as the independent variable, outness to variable. The main effect for locale was not significant: peers as a covariate, and interactions between locale F(2, 8113)=1.002, p>0.10. and outness. The main effect for locale was significant: F(2, 7953)=3.938, p<0.05. The interaction effect 84 To test differences across locale in comfort talking to was not significant. Post-hoc Bonferroni tests indicated school staff about LGBT issues, a multivariate analysis of that rural students reported lower levels of self-esteem variance (MANOVA) was conducted with comfort talking than suburban students. Suburban students were not with school staff about LGBT issues as the dependent statistically different from urban students. variables and locale as the independent variable. Multivariate results were significant: Pilai’s trace=.007, 78 Depression was measured using the 20-item Likert-type F(3, 7648)=3.138, p<.001. Univariate effects were CES-D depression scale (Eaton et al., 2004), which significant atp <.01, except for nurses, who were includes such items as “During the past week, I felt significant at thep <.05 level. hopeful about the future.” 85 Based on a repeated-measures multivariate analysis of Eaton, W. W., Muntaner, C., Smith, C., Tien, A., & variance (MANOVA): students were significantly more Ybarra, M. (2004). Center for Epidemiologic Studies comfortable talking with teachers and counselors than Depression Scale: Review and revision (CESD and with other types of staff members. Pilai’s trace=.509, CESD-R) (pp. 363-377). In M. E. Maruish (Ed.), The F(21, 53536)=1131.359, p<.001. Univariate effects use of psychological testing for treatment planning and were significant at thep <.001 level. outcomes assessment. 3rd Ed. Mahwah, NJ: Lawrence Erlbaum. 86 To test differences in discussing LGBT issues with teachers by locale, an analysis of variance (ANOVA) was conducted with frequency of discussing LGBT issues as the dependent variable and locale as the independent variable. The main effect for locale was significant: F(2, 7775)=7.995, p<0.01.

Strengths and Silences 39 87 To test differences in discussing LGBT issues with 93 Having an inclusive curriculum was included in the counselors by locale, an analysis of variance (ANOVA) multiple analysis of variance (MANOVA) conducted was conducted with frequency of discussing LGBT issues above, with presence of an inclusive curriculum as as the dependent variable and locale as the independent a dependent variable and locale as the independent variable. The main effect for locale was significant: F(2, variable. The main effect of locale was significant: F(2, 7775)=3.736, p<0.05. 7905)=27.374, p<0.001, effect size =.007. Post- hoc Bonferroni tests showed that rural students were 88 To test differences in discussing LGBT issues with significantly less likely to attend schools with inclusive safety/resource officers by locale, an analysis of variance textbooks than were suburban or urban students. (ANOVA) was conducted with frequency of discussing LGBT issues as the dependent variable and locale as 94 Having Internet access to LGBT resources was included the independent variable. The main effect for locale was in the multiple analysis of variance (MANOVA) conducted significant: F(2, 7775)=4.697,p <0.01. above, with Internet access to LGBT resources as a dependent variable and locale as the independent 89 To test differences in school belonging by locale and variable. The main effect of locale was significant: F(2, frequency of talking to teachers, an analysis of variance 7905)=8.944, p<0.001, effect size =.002. Post- (ANOVA) was conducted with school belonging as the hoc Bonferroni tests showed that rural students were dependent variable, locale and frequency of talking to significantly less likely to attend schools with internet teachers as the independent variables, and interaction access to LGBT resources than were suburban or urban effects between locale and frequency of talking to students. a teacher. The effect of locale was significant: F(2, 8079)=57.583, p<0.001. The main effect of talking 95 Having access to LGBT library resources was included in to a teacher about LGBT issues was significant: the multiple analysis of variance (MANOVA) conducted F(1, 8079)=424.003, p<0.001. The interaction above, with access to LGBT library resources as a effectsbetween locale and talking to a teacher was not dependent variable and locale as the independent significant. variable. The main effect of locale was not significant: F(2, 7905)=.916, p>0.10. 90 To test differences in school belonging across frequency of talking to school personnel, several analyses of 96 Heitel Yakush, J. (2007) Legalized discrimination: variance (ANOVA) were conducted with the mean The rise of the marriage-promotion industry and how score on the school belonging scale as the dependent federally funded programs discriminate against lesbian, variable, and frequency of talking to a school staff gay, bisexual, and transgender youth and families. person about LGBT issues as the independent variable. Washington, D.C., Sexuality Information and Education The main effect of frequency of talking to a teacher Council of the United States (SIECUS). Available at: was significant: F(1, 2369)=109.159,p <0.001, effect http://www.siecus.org/_data/global/images/Legalized- size=.044. The main effect of frequency of talking to a Discrimination.pdf. principal was significant: F(1, 2361)=33.655,p <0.001, effect size=.014. The main effect of frequency of talking Ott, M. & Santelli, J. (2007). Abstinence and to a vice principal was significant: F(1, 2344)=27.324, abstinence-only education. Current Opinion in Obstetrics p<0.001, effect size=.012. The main effect of & Gynecology, 19(5), 446–452. frequency of talking to a counselor was significant: F(1, 2355)=15.489, p<0.001, effect size= .007. Santelli J., Ott, M., Lyon, M., Roger, J., Summer, D., & The main effect of talking to a nurse was significant: Schleifer, R. (2006) Abstinence and abstinence-only F(1, 2354)=8.880, p<0.01, effect size=.004. The education: A review of US policies and programs. Journal main effect of frequency of talking to a librarian was of Adolescent Health, 38(1), 72–81. significant: F(1, 2355)=27.581,p <0.001, effect size=.012. The main effect of frequency of talking to Sexuality Information and Education Council of the a gym teacher was significant: F(1, 2352)=20.011, United States (SIECUS) (2004). A portrait of sexuality p<.001, effect size=.008. The main effect of frequency education and abstinence-only-until-marriage programs of talking to a safety/ resource officer was not significant: in the states. Available at: http://www.siecus.org/index. F(1, 2352)=1.091, p>.10. cfm?fuseaction=Page.viewPage&pageId=487&parent ID=478. 91 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, Madelyn J., & Palmer, N. A. (2012). The 2011 National U.S. House of Representatives Committee on School Climate Survey: The experiences of lesbian, gay, Government Reform – Minority Staff Special bisexual, and transgender youth in our nation’s schools. Investigations Division. (December 2004). The content New York: Gay, Lesbian & Straight Education Network. of federally funded abstinence-only education programs, prepared for Rep. Henry A. Waxman. 92 To test differences in access to LGBT-related resources and supports across locales, a multivariate analysis of Sexuality Information and Education Council of the variance (MANOVA) was conducted with resources as United States (SIECUS) (2004). A portrait of sexuality the dependent variables and locale as the independent education and abstinence-only-until-marriage programs variable. The multivariate effect was significant: in the states. Available at: http://www.siecus.org/index. Pillai’s trace=.92, F(18, 15816)=42.150, p<.001. cfm?fuseaction=Page.viewPage&pageId=487&parent The univariate effect of locale in having an inclusive ID=478. curriculum was significant: F(2, 7905)=37.033, p<0.01, effect size=.009. Post-hoc Bonferroni tests showed that rural students were significantly less likely to attend schools with inclusive curricula than were suburban or urban students.

40 Strengths and Silences 97 A chi-square test was conducted to compare the 104 To test differences in access to a community group percentages of students who reported being taught about or program by locale and school size, an analysis of sexual health by locale: χ2=18.716, df=2, p<.001, variance (ANOVA) was conducted with access to a Ф=.048. Adjusted p-values indicated that rural and community group/program as the dependent variable, urban students were less likely to have been taught locale and the presence of a GSA as independent about sexual health than suburban students. variables, and interaction terms between locale and GSA presence. The main effect for locale was significant: 98 A chi-square test was conducted to compare the F(2, 8111)=43.663, p<0.001. The main effect for percentages of students who reported being taught an GSA presence was significant: F(1, 8111)=568.899, abstinence-only curriculum by locale: χ2=29.277, df=2, p<0.001. The interaction term was also significant: p<.001, Ф=.060. Adjusted p-values indicated that F(2, 8111)=9.870, p<0.001, with suburban students rural students were more likely to have been taught an more likely than rural students to have a community abstinence-only curriculum than suburban and urban group even when they lacked access to a GSA (with students. significance at thep <.001 level).

99 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, 105 Klem, A. M., & Connell, J. P. (2004). Relationships Madelyn J., & Palmer, N. A. (2012). The 2011 National matter: Linking teacher support to student engagement School Climate Survey: The experiences of lesbian, gay, and achievement. Journal of School Health, 74(7), 262- bisexual, and transgender youth in our nation’s schools. 273. New York: Gay, Lesbian & Straight Education Network. 106 Number of school staff supportive of LGBT students was 100 Having a GSA was included in the multiple analysis included in the multiple analysis of variance (MANOVA) of variance (MANOVA) conducted above, with havign conducted above, with number of supportive staff as a GSA at school as a dependent variable and locale a dependent variable and locale as the independent as the independent variable. The main effect of locale variable. The main effect of locale was significant: F(2, was significant: F(2, 7905)=259.247,p <0.001, effect 7905)=134.737, p<0.001, effect size=.033. Post- size =.062. Post-hoc Bonferroni tests showed that rural hoc Bonferroni tests showed that rural students had students were significantly less likely to attend schools significantly fewer supportive staff than suburban or with GSAs than were suburban or urban students. urban students.

101 To test differences in GSA attendance across locale, an 107 To test differences across locale in having a school analysis of variance (ANOVA) was conducted with the administration supportive of LGBT students, an analysis “frequency of GSA attendance as the dependent variable of variance (ANOVA) was conducted with supportiveness and locale as the independent variable. The main effect of administration as the dependent variable, and of locale was significant: F(2, 3709)=7.242,p <0.01, locale as the independent variable. The main effect of effect size=.004. locale was significant: F(2, 8044)=64.865,p <0.001. The effect size was .016. Post-hoc Bonferroni tests 102 To test this possibility, an analysis of covariance showed that rural students rated their administration as (ANCOVA) was conducted with the frequency of significantly less supportive than suburban and urban attending a GSA as the dependent variable, locale as the students. independent variable, frequency of hearing homophobic remarks as a covariate, and interaction terms between 108 Anhalt, K., & Morris, T. L. (1998). Developmental locale and frequency of hearing homophobic remarks. and adjustment issues of gay, lesbian, and bisexual The overall interaction effect just missed marginal adolescents: A review of the empirical literature. Clinical significance, atp <.11. The interaction effects between Child and Family Psychology Review, 1(4), 215-230. locale and homophobic remarks were significant for rural compared to suburban (t=2.082, p<.05) and marginally D’Augelli, A. R., & Hershberger, S. L. (1993). Lesbian, significant for rural compared to urban students gay, and bisexual youth in community settings: Personal (t=1.764, p<0.10). challenges and mental health problems. American Journal of Community Psychology, 21(4), 421-448. 103 Having a community group or program for LGBT youth was included in the multiple analysis of variance Diamond, L. M., & Lucas, S. (2004). Sexual-minority (MANOVA) conducted above, with presence of a group and heterosexual youths’ peer relationships: Experiences, or program as a dependent variable and locale as the expectations, and implications for well-being. Journal of independent variable. The main effect of locale was Research on Adolescence, 14(3), 313-340. significant: F(2, 7905)=119.482,p <0.001, effect size =.029. Post-hoc Bonferroni tests showed that 109 To test differences across locale in peer acceptance rural students were significantly less likely to live in of LGBT students, an analysis of variance (ANOVA) communities with a group or program for LGBT youth was conducted with acceptance of peers as the than were suburban or urban students. dependent variable and locale as the independent variable. The main effect of locale was significant: F(2, 8130)=146.004, p<0.001, effect size=.035. Post- hoc Bonferroni tests showed that rural students rated their peers as less accepting than suburban andurban students.

Strengths and Silences 41 110 Smith, J. D. (1997). Working with larger systems: Rural 117 To test differences in victimization based on sexual lesbians and gays. In J. D. Smith & R. J. Mancoske orientation for rural LGBT students by presence of (Eds.), Rural gays and lesbians: Building on the strength supportive resources, several analyses of variance of communities (pp. 13-22). New York, NY: The (ANOVAs) were conducted, with weighted victimization Harrington Park Press. score as the dependent variable and supportive resource as the independent variable. The main effect for Swank, E. D., Frost, D. M., & Fahs, B. (2012). Rural having many supportive educators was significant: F(1, location and exposure to minority stress among sexual 1110)=43.311, p<0.001, effect size=.036. The main minorities in the United States. Psychology & Sexuality, effect for an inclusive curriculum was significant: F(1, 3(3), 226-243. 2349)=52.129, p<0.001, effect size=.022. The main effect for having a comprehensive policy relative to no 111 To test differences in number of LGBT peers across policy was significant: F(1, 553)=3.626,p <0.06, effect locale, an analysis of covariance (ANCOVA) was size=.007. conducted with number LGBT peers as the dependent variable (weighted to the approximate value of the 118 To test differences in victimization based on gender response option in the ordinal response set), locale expression for rural LGBT students by presence of as the independent variable, and school size as a supportive resources, several analyses of variance covariate. The main effect of locale was significant: F(2, (ANOVAs) were conducted, with weighted victimization 8074)=19.303, p<0.001. Bonferroni post-hoc tests score as the dependent variable and supportive resource indicated that urban students knew a greater number as the independent variable. The main effect for the of LGBT peers than rural and suburban students. Rural presence of a GSA was significant: F(1, 2296)=20.274, students knew fewer LGBT peers than suburban students p<0.001, effect size=.009. The main effect for having at marginal significance:p <0.10. many supportive educators was significant: F(1, 1083)=23.158, p<0.001, effect size=.021. The main 112 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, effect for an inclusive curriculum was significant: F(1, Madelyn J., & Palmer, N. A. (2012). The 2011 National 2295)=30.080, p<0.001, effect size=.013. The main School Climate Survey: The experiences of lesbian, gay, effect for having a comprehensive policy relative to no bisexual, and transgender youth in our nation’s schools. policy was not significant: F(1, 546)=3.342,p >.10, New York: Gay, Lesbian & Straight Education Network. effect size=.004.

113 A chi-square test was conducted to compare the 119 To test differences in school belonging for rural LGBT percentages of students who reported having a students by presence of many supportive teachers versus comprehensive policy in their school by locale: none, an analysis of variance (ANOVA) was conducted χ2=34.492, df=2, p<.001, Ф=.065. Adjusted p-values with the school belonging score as the dependent indicated that rural students were less likely to have variable, and presence of many supportive educators as comprehensive policies than suburban students, who the independent variable. The main effect for supportive were less likely to have them than urban students. educators was significant: F(1, 1119)=206.339, p<0.001, effect size=.156. 114 A chi-square test was conducted to compare the percentages of students who reported having any policy 120 To test differences in school belonging for rural LGBT in their school by locale: χ2=18.104, df=2, p<.001, students by presence of a GSA, an analysis of variance Ф=.047. Adjusted p-values indicated that rural and (ANOVA) was conducted with the school belonging score suburban students were less likely to have an anti- as the dependent variable, and presence of a GSA as the bullying and harassment policy than urban students. independent variable. The main effect for the presence of a GSA was significant: F(1, 2371)=102.607, 115 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, p<0.001, effect size=.041. To test differences in school Madelyn J., & Palmer, N. A. (2012). The 2011 National belonging for rural LGBT students by presence of an School Climate Survey: The experiences of lesbian, gay, inclusive curriculum, an analysis of variance (ANOVA) bisexual, and transgender youth in our nation’s schools. was conducted with the school belonging score as New York: Gay, Lesbian & Straight Education Network. the dependent variable, and presence of an inclusive curriculum as the independent variable. The main 116 To test differences in victimization based on sexual effect for inclusive curriculum was significant: F(1, orientation for rural LGBT students by presence of a 2371)=205.402, p<0.001, effect size=.080. To test GSA, an analysis of variance (ANOVA) was conducted differences in school belonging for rural LGBT students with the weighted victimization score as the dependent by presence of a comprehensive policy versus no variable, and presence of a GSA as the independent policy, an analysis of variance (ANOVA) was conducted variable. The main effect for the presence of a GSA with the school belonging score as the dependent was significant: F(1, 2349)=43.044,p <0.001, effect variable, and presence of a comprehensive policy as the size=.018. independent variable. The main effect for comprehensive was significant: F(1, 558)=58.460,p <.001, effect size=.095.

42 Strengths and Silences 121 To test differences in depression for rural LGBT students 126 Hispanic/Latino and Middle Eastern/Arab American by presence of supportive resources, several analyses of categories were considered ethnicities as opposed variance (ANOVAs) were conducted, with the depression to races, and thus students selecting either of those score as the dependent variable and supportive resource categories were coded as such, regardless of race (e.g., as the independent variable. The main effect for the student selecting “African-American” and “Latino/a” presence of a GSA was significant: F(1, 2329)=33.135, were coded as “Latino/a”). p<0.001, effect size=.014. The main effect for having many supportive educators was significant: F(1, When forced to select one response, students with both 1100)=87.326, p<0.001, effect size=.074. The main White and another racial background may be more effect for an inclusive curriculum was significant: F(1, likely to select a non-White identity, particularly when 2329)=88.371, p<0.001, effect size=.037. The main “multiracial” is not an option: effect for having a comprehensive policy relative to no policy was significant: F(1, 550)=9.183,p <.01, effect Aud, S., Fox, M., & KewalRamani, A. (2010). Status size=.016. and trends in the education of racial and ethnic groups (NCES 2010-015). U.S. Department of Education, 122 To test differences in self-esteem for rural LGBT students National Center for Education Statistics. Washington, by presence of supportive resources, several analyses of DC: U.S. Government Printing Office. variance (ANOVAs) were conducted, with the self-esteem score as the dependent variable and supportive resource This method may result in a higher percentage of as the independent variable. The main effect for the students of color from specific racial groups being presence of a GSA was significant: F(1, 2341)=5.420, identified in other surveys and a higher percentage of p<0.001, effect size=.005. The main effect for having students being identified as multiracial in our survey. many supportive educators was significant: F(1, This difference in method may account for some of the 1104)=71.380, p<0.001, effect size=.0616. The main discrepancy regarding percentages of specific racial effect for an inclusive cur-riculum was significant: F(1, groups (e.g., African American/Black, Asian/ Pacific 2341)=58.253, p<0.001, effect size=.024. The main Islander) between our sample and the general population effect for having a comprehensive policy relative to no of secondary school students. policy was significant: F(1, 549)=14.166,p <.001, effect size=.025. 127 Gray, M. L. (2009). Negotiating identities/queering desires: Coming out online and the remediation of 123 A one-sample chi-square test was conducted to compare the coming-out story. Journal of Computer-Mediated the locale distribution of students in the 2011 NSCS to Communication, 14, 1162-1189. the locale distribution of primary and secondary school students reported for 2010-2011 by the National Center Gray, M. L. (2009). Out in the country: Youth, media, for Education Statistics (NCES): χ2=262.063, df=2, and queer visibility in rural America. New York: New York p<.001, Ф=.175. University Press.

National Center for Education Statistics (2011). NCES Herring, S. (2010). Another country: Queer anti- Common Core of Data (CCD). Public elementary/ urbanism. New York: New York University Press. secondary school universe survey: School year 2010- 2011, Final Version 1a. Retrieved from http://nces. 128 Kosciw, J. G., Greytak, E. A., Bartkiewicz, M. J., Boesen, ed.gov/ pubs2012/pesschools10/tables/table_06.asp. Madelyn J., & Palmer, N. A. (2012). The 2011 National School Climate Survey: The experiences of lesbian, gay, 124 National Center for Education Statistics (2010). NCES bisexual, and transgender youth in our nation’s schools. Common Core of Data (CCD). Public elementary/ New York: Gay, Lesbian & Straight Education Network. secondary school universe survey: School year 2007–08. Retrieved from http://nces.ed.gov/pubs2010/2010015/ tables/table_7_2.asp.

Provasnik, S., KewalRamani, A., Coleman, M. M., Gilbertson, L., Herring, W., & Xie, Q. (2007). Status of education in rural America (NCES 2007-040). U.S. Department of Education, National Center for Education Statistics. Washington, DC: U.S. Government Printing Office.

125 Herman, M. (2004). Forced to choose: Some determinants of racial identification in multiracial adolescents. Child Development, 75(3), 730–748.

Strengths and Silences 43

90 Broad Street, 2nd Floor New York, NY 10004 Ph: 212.727.0135 Fax: 212.727.0254 www.glsen.org 48 Strengths and Silences